A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence
Open accessAbstract
A number of algorithms in the field of artificial intelligence offer poorly interpretable decisions. To disclose the reasoning behind such algorithms, their output can be explained by means of so-called evidence-based (or factual) explanations. Alternatively, contrastive and counterfactual explanations justify why the output of the algorithms is not any different and how it could be changed, respectively. It is of crucial importance to bridge the gap between theoretical approaches to contrastive and counterfactual explanation and the corresponding computational frameworks. In this work we conduct a systematic literature review which provides readers with a thorough and reproducible analysis of the interdisciplinary research field under study. We first examine theoretical foundations of contrastive and counterfactual accounts of explanation. Then, we report the state-of-the-art computational frameworks for contrastive and counterfactual explanation generation. In addition, we analyze how grounded such frameworks are on the insights from the inspected theoretical approaches. As a result, we highlight a variety of properties of the approaches under study and reveal a number of shortcomings thereof. Moreover, we define a taxonomy regarding both theoretical and practical approaches to contrastive and counterfactual explanation.
AI snapshot
Written from the paper’s full text.
TL;DR
This systematic literature review maps theoretical accounts and computational frameworks for contrastive and counterfactual explanation in XAI, proposing the umbrella term “contfactual explanation.” It reports a predominantly causal theoretical literature, many task-specific generation frameworks, and a weak link between theory and implementation, with evaluation methods lacking standardization.
Key findings
- From 113 unique primary studies, 74/113 (65.49%) addressed definitions/theoretical accounts (RQ1), 52/113 (46.02%) described generation frameworks (RQ2), and only 13/113 (11.50%) addressed both (RQ3).
- Among theoretical accounts, 49/74 (66.22%) concerned causal explanation; the proposed taxonomy distinguishes causal, non-causal, and hybrid contfactual explanation.
- Among 52 generation frameworks, 31 (59.62%) generated counterfactual explanations, 17 (32.69%) contrastive, and 4 (7.69%) fused both; 42/52 (80.77%) targeted classification. Most were model-agnostic counterfactual methods, and most output numerical/feature-based explanations.
- Evaluation was non-standardized: 17/52 (32.69%) provided no evaluation details; most used objective metrics, few used hybrid automatic+human evaluation. Only 13/113 studies were theoretically grounded, and only 3/13 (23.08%) of those interpreted theory to propose their own definition.
Methods
- Systematic literature review following Kitchenham and Charters, Kitchenham et al., and Wohlin, with planning, conducting, and reporting phases; three RQs on definitions, generation methods, and theory-practice grounding.
- Searched Scopus and Web of Science on October 2, 2020 using six queries (q1–q6) over title, abstract, and author keywords with word stems and NEAR/3/W proximity operators; applied inclusion/exclusion criteria, duplicate removal, and snowballing.
- From 338 candidate primary studies, 136 duplicates removed, 75 filtered as irrelevant, 33 discarded for inclusion criteria, and 19 added via bibliography inspection, yielding 113 unique primary studies; data extraction covered definitions/theories and framework problem, method, output form, and evaluation, plus bibliometric analysis with Mendeley and VOSViewer.
Limitations
- The paper states that no uniform set of evaluation methods has been adopted, that many frameworks lack standardization, and that a gap remains between philosophical accounts and ML-related concepts, making theoretical frameworks poorly applicable to XAI.
- The paper states that definitions are highly task- and domain-specific, that most computational frameworks are non-causal and not rooted in theoretical accounts, and that only 13/113 studies were theoretically grounded.
- Evident from the text: the search was limited to Scopus and WoS and to a single date (October 2, 2020), so 2020 coverage is incomplete; the review also notes that many frameworks are limited to specific models, tasks, or small feature sets, and that user-centered evaluation is rare.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 220 cited sources checked · 0 supported · 0 flagged
- Findingnone given
A major factor discerning theoretical approaches to counterfactual explanation is their relation to causation.
“However, a major factor discerning theoretical approaches to contfactual explanation is found to consist in their relation to causation.”
Quote in the paperSubstantial claim - Findingown data
There was a rapid rise in the number of publications in the past three years.
“Furthermore, Fig. 5 shows a rapid rise in the number of publications in the past three years.”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
- Findingown data
Causal counterfactual makes up a majority of the considered theoretical frameworks.
“causal counterfactual making up a majority of the considered theoretical frameworks (see Fig. 8).”
Quote in the paperNot a substantial claim- Rests on an assumption: Counterfactual explanation encompasses contrastive, counterfactual, and contrastive-counterfactual explanation.
- Findingown data
Each type of counterfactual explanation is present in the findings.
“Each type of contfactual explanation is present in the findings”
Quote in the paperSubstantial claim- Rests on an assumption: Counterfactual explanation encompasses contrastive, counterfactual, and contrastive-counterfactual explanation.
- Findingown data
Most counterfactual explanation generation frameworks deal with counterfactual explanation (31 out of 52 frameworks; 59.62%).
“As depicted in Fig. 9, most contfactual explanation generation frameworks deal with counterfactual explanation (31 out of 52 frameworks; 59.62%).”
Quote in the paperNot a substantial claim
- Findingown data
13 out of 113 (11.50%) primary studies address both RQ 1 and RQ 2 and therefore answer RQ 3.
“Note that 13 out of 113 (11.50%) primary studies were found to address both RQ 1 and RQ 2 and therefore answer RQ 3 .”
Quote in the paperNot a substantial claim - Findingcitation
One of the studies that fuse contrastive and counterfactual explanations [129] deals with both classification and regression.
“One of these studies [129] deals with both classification and regression.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
- Findingown data
Only four studies (7.69%) fuse contrastive and counterfactual explanations.
“Only four studies (7.69%) fuse contrastive and counterfactual explanations.”
Quote in the paperNot a substantial claim - Findingown data
74 out of 113 (65.49%) primary studies formulate definitions for counterfactual explanation and/or address theoretical accounts thereof (RQ 1).
“Thus, a collection of 74 out of 113 (65.49%) original primary studies were found to formulate definitions for contfactual explanation and/or address theoretical accounts thereof (RQ 1 ).”
Quote in the paperNot a substantial claim - Findingown data
The outlined classification-oriented frameworks are evaluated on classifiers based on logistic regression, decision trees, gradient boosted decision trees, support vector machines, random forests, neural networks, or combinations of these.
“Thus, the outlined classification-oriented frameworks are evaluated on classifiers based on logistic regression [55], [136], [153], [158], decision trees [46], [80], [122], [140], [150], [155], [159], gradient boosted decision trees [147], support vector machines [131], [138], [146], random forests [81], [86], [142]-[144], neural networks [6], [48], [49], [91], [129], [130], [133], [135], [139], [”
Quote in the paperNot a substantial claim0 of 3 sources checked - Findingreasoning
There have been several attempts to measure a degree of causality in counterfactual explanation.
“There have been several attempts to measure a degree of causality in contfactual explanation.”
Quote in the paperNot a substantial claim - Findingcitation
A large body of research on both causal and non-causal counterfactual explanation testifies that counterfactuals have a diverse nature with respect to causality.
“Indeed, a large body of research on both causal and non-causal counterfactual explanation testifies that counterfactuals have a diverse nature with respect to causality [104].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [104] source
Not checked in the free run.
- Findingown data
A sum of 338 publications (207 from Scopus and 131 from Web of Science) made up the collection of candidate primary studies addressing the research questions.
“A sum of 338 publications (207 from Scopus and 131 from WoS) made up the collection of CPS addressing the research questions.”
Quote in the paperSubstantial claim - Findingnone given
There have been several attempts to unify causal and non-causal counterfactuals within one framework.
“In light of this, there have been several attempts to unify causal and non-causal counterfactuals within one framework.”
Quote in the paperNot a substantial claim - Findingown data
Most considered studies on causal counterfactual explanation relate to either of four theoretical milestones: Lewis-Stalnaker's theory of closest possible worlds, Pearl's Structural Causal Models, Woodward's Counterfactual Theory of Explanation, or the Neyman-Rubin Causal Model.
“Most of the considered studies on causal counterfactual explanation relate to either of the four theoretical milestones: Lewis-Stalnaker's theory of closest possible worlds [36], [120], Pearl's Structural Causal Models (SCM) [37], Woodward's Counterfactual Theory of Explanation (CTE) [125], or the Neyman-Rubin Causal Model (NRCM) [107], [115] (see Table 5).”
Quote in the paperNot a substantial claim0 of 3 sources checked - Findingown data
Counterfactual explanation appears to attract increasing attention across all subject areas in the past two decades.
“It can be seen that contfactual explanation appears to attract an increasing attention across all subject areas in the past two decades.”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
- Findingown data
17 out of 52 (32.69%) generate contrastive explanations.
“In contrast, 17 out of 52 (32.69%) generate contrastive explanations.”
Quote in the paperNot a substantial claim - Findingnone given
Several accounts of counterfactual explanation presuppose a causal nature and establish a causal counterfactual dependency between the phenomenon to be explained and the explanation itself.
“As will be shown below, several accounts of contfactual explanation presuppose a causal nature and establish a causal contfactual dependency between the phenomenon to be explained and the explanation itself.”
Quote in the paperNot a substantial claim - Findingown data
A vast majority of state-of-the-art AI applications that generate counterfactuals (42 out of 52; 80.77%) are used to explain the outcome of ML-based classifiers.
“A vast majority of state-of-the-art AI applications that generate contfactuals (42 out of 52; 80.77%) are used to explain the outcome of ML-based classifiers, i.e., algorithms that learn a mapping function f : X -→ Y from a training dataset of n labeled examples X = {x i | 1 ≤ i ≤ n} to a discrete output variable (class) Y = {y j | 1 ≤ j ≤ m} where m is the number of classes.”
Quote in the paperNot a substantial claim - Findingreasoning
113 unique publications formed the exhaustive pool of primary studies.
“As a result, 113 unique publications formed the exhaustive pool of primary studies.”
Quote in the paperSubstantial claim - Findingcitation
Non-causal contrastive explanations can address the physical nature of a modeled system.
“Notably, noncausal contrastive explanations can address the physical nature of a modeled system.”
Quote in the paperNot a substantial claim - Findingown data
Counterfactual explanation is often investigated in the context of causation, cognitive science, and AI.
“It can be concluded that counterfactual explanation is often investigated in the context of causation (pay attention to such keywords as ''causation'', ''causal inference'' or ''causal models'') as well as cognitive science (as reflected by the keywords ''imagination'', ''reasoning'', etc.) and AI (''machine learning'', ''data models'', ''black-box models'').”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
- Findingown data
52 out of 113 (46.02%) publications describe frameworks or extensions of other frameworks for counterfactual explanation generation (RQ 2).
“In addition, 52 out of 113 (46.02%) publications describe frameworks (or extensions of other frameworks) for contfactual explanation generation (RQ 2 ).”
Quote in the paperNot a substantial claim - Findingcitation
Two model-specific frameworks are found in the context of video processing.
“Two model-specific frameworks are found in the context of video processing.”
Quote in the paperNot a substantial claim - Findingcitation
Several model-specific frameworks generate counterfactuals to explain the output of decision trees [46], [80], [122], [155].
“Several model-specific frameworks generate counterfactuals to explain the output of decision trees [46], [80], [122], [155].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Findingown data
Most frameworks deal with counterfactual model-agnostic methods.
“Most frameworks deal with counterfactual model-agnostic methods.”
Quote in the paperNot a substantial claim - Findingnone given
A large number of frameworks are limited to explaining the output of particular models due to task-specific constraints.
“A large number of frameworks are limited to explaining the output of particular models due to task-specific constraints.”
Quote in the paperNot a substantial claim - Findingnone given
A number of frameworks address specific properties of counterfactuals.
“A number of frameworks address specific properties of counterfactuals.”
Quote in the paperNot a substantial claim - Findingreasoning
Most explanation generators meant for planning-based tasks are model-specific due to the problem-and approach-specific restrictions preventing them from being used for other AI challenges.
“Most explanation generators meant for planning-based tasks are model-specific due to the problem-and approachspecific restrictions preventing them from being used for other AI challenges.”
Quote in the paperNot a substantial claim - Findingnone given
Several model-agnostic frameworks aim at measuring specific properties of counterfactuals.
“Several model-agnostic frameworks aim at measuring specific properties of contfactuals.”
Quote in the paperNot a substantial claim - Findingcitation
The problem of contrastive explanation generation for planning is also found to be framed in the reinforcement learning setting.
“The problem of contrastive explanation generation for planning is also found to be framed in the reinforcement learning setting.”
Quote in the paperNot a substantial claim - Findingnone given
A large number of model-agnostic frameworks treat counterfactual explanation generation as an optimization problem in a post-hoc manner.
“A large number of model-agnostic frameworks treat contfactual explanation generation as an optimization problem in a post-hoc manner.”
Quote in the paperNot a substantial claim - Findingown data
As depicted in Fig. 11, most frameworks focus on numerical counterfactual output.
“As depicted in Fig. 11, most frameworks focus on numerical counterfactual output.”
Quote in the paperNot a substantial claim - Findingown data
The considered frameworks output counterfactual explanations in several ways.
“The considered frameworks output contfactual explanations in several ways.”
Quote in the paperNot a substantial claim - Findingnone given
Similarly to the model-specific frameworks, numerous model-agnostic explanation generators are found to be task-specific.
“Similarly to the model-specific frameworks, numerous model-agnostic explanation generators are found to be task-specific.”
Quote in the paperNot a substantial claim - Findingcitation
A number of the frameworks are evaluated by means of accuracy-based metrics.
“A number of the frameworks are evaluated by means of accuracy-based metrics [61], [91], [94], [159].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Findingreasoning
A strikingly low number of computational frameworks appear to be rooted in theoretical accounts of explanation due to an imbalance in favor of causality-oriented theoretical accounts.
“Furthermore, a strikingly low number of such frameworks appear to be rooted in theoretical accounts of explanation due to an imbalance in favor of causality-oriented theoretical accounts.”
Quote in the paperNot a substantial claim - Findingnone given
All the considered frameworks are evaluated by means of intrinsic (either subjective or objective) metrics.
“It is worth noting that all the considered frameworks are evaluated by means of intrinsic (either subjective or objective) metrics.”
Quote in the paperSubstantial claim - Findingown data
A majority of the considered theoretical accounts of counterfactual explanation (49 out of 74; 66.22%) speculate on the causal nature of explanation.
“The results of the study, in relation to RQ 1 , show that a majority of the considered theoretical accounts of contfactual explanation (49 out of 74; 66.22%) speculate on the causal nature of explanation.”
Quote in the paperSubstantial claim - Findingcitation
An increasing number of interactive frameworks attempt not only to present the automatically generated explanations to the end-user but also interact with him or her.
“However, we find an increasing number of interactive frameworks that attempt not only to present the automatically generated explanations to the end-user but also interact with him or her [46], [150], [155].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Findingnone given
The two aforementioned studies [128], [137] lack an evaluation of the quality of the generated counterfactual explanations themselves.
“However, the two aforementioned studies [128], [137] lack an evaluation of the quality of the generated contfactual explanations themselves.”
Quote in the paperNot a substantial claim0 of 2 sources checked - Findingown data
17 out of 52 (32.69%) of the considered publications do not evaluate their frameworks.
“17 out of 52 (32.69%) of the considered publications do not evaluate their frameworks, i.e., neither automatic metrics for contfactual explanation generation are suggested nor a human evaluation survey is presented in such publications.”
Quote in the paperNot a substantial claim - Findingnone given
The other computational frameworks are primarily non-causal.
“At the same time, the other computational frameworks are primarily non-causal.”
Quote in the paperNot a substantial claim - Findingown data
Only 13 out of 113 studies (11.50%) were present in both of the pools of primary studies related to RQ1 and RQ2.
“Indeed, only 13 out of 113 studies (11.50%) were present in both of the pools of primary studies related to RQ 1 and RQ 2 .”
Quote in the paperSubstantial claim - Findingnone given
Despite an increasingly expanding use of counterfactual explanations, no uniform set of evaluation methods has been adopted so far.
“Unfortunately, despite an increasingly expanding use of contfactual explanations, no uniform set of evaluation methods has been adopted so far.”
Quote in the paperSubstantial claim - Findingcitation
An extensive number of evaluation methods are found to be strictly task- or approach-specific.
“An extensive number of evaluation methods are found to be strictly task-or approach-specific.”
Quote in the paperNot a substantial claim - Findingown data
A large body of research has been elaborated on theoretical accounts of contrastive, counterfactual, and contrastive-counterfactual explanation.
“The findings show that a large body of research has been elaborated on theoretical accounts of contrastive, counterfactual, and contrastive-counterfactual explanation.”
Quote in the paperNot a substantial claim - Findingown data
50 out of the 52 considered state-of-the-art counterfactual explanation generation frameworks (96.15%; RQ2) have been developed from 2017 to 2020.
“Thus, 50 out of the 52 considered state-of-the-art contfactual explanation generation frameworks (96.15%;RQ2) have been developed from 2017 to 2020.”
Quote in the paperNot a substantial claim - Findingreasoning
Most of the modern counterfactual explanation generators focus on feature perturbation when searching for the most relevant counterfactuals and not establishing causal relations between them.
“This partly explains why most of the modern contfactual explanation generators focus on feature perturbation when searching for the most relevant contfactuals and not establishing causal relations between them.”
Quote in the paperNot a substantial claim - Findingnone given
Counterfactuals are found to address both causal and non-causal dependencies.
“Contfactuals are found to address both causal and non-causal dependencies.”
Quote in the paperNot a substantial claim - Findingown data
Only 3 out of the 13 (23.08%) studies interpret the insights from the theoretical foundations to propose their own counterfactual explanation definition for problem-oriented purposes.
“Moreover, only 3 out of the 13 (23.08%)studies interpret the insights from the theoretical foundations to propose their own contfactual explanation definition for problem-oriented purposes.”
Quote in the paperSubstantial claim - Findingreasoning
Other considered frameworks do not appeal directly to any of the theoretical accounts of explanation addressed in RQ1.
“Instead, other considered frameworks do not appeal directly to any of the theoretical accounts of explanation addressed in RQ 1 .”
Quote in the paperNot a substantial claim - Findingreasoning
A crucial shortcoming relevant to the inspected frameworks is a lack of standardization with respect to the evaluation methods.
“A crucial shortcoming relevant to the inspected frameworks is a lack of standardization with respect to the evaluation methods.”
Quote in the paperSubstantial claim - Findingown data
Most frameworks deal with objective evaluation of counterfactuals.
“It is easy to appreciate how most frameworks deal with objective evaluation of counterfactuals.”
- Also stated in: § FIGURE 12. · plain
“A majority of the researchers propose objective (automatic) methods for evaluating automatically generated contfactuals.”
Quote in the paperNot a substantial claim - Also stated in: § FIGURE 12. · plain
- Findingown data
Only few of the existing computational frameworks are grounded on theories of counterfactual explanation.
“We find that only few of the existing computational frameworks are grounded on theories of contfactual explanation.”
Quote in the paperSubstantial claim - Findingcitation
Redefined contrastive explanations are also found in the domain of robotics and automatic planning.
“Redefined contrastive explanations are also found in the domain of robotics and automatic planning.”
Quote in the paperNot a substantial claim - Premisecitation
Lombrozo treated counterfactual and contrastive explanations as equivalent assuming hypothesized events non-occurred in reality to be counterfactual cases where a subset of these cases forms a contrastive explanation.
“For instance, Lombrozo treated counterfactual and contrastive explanations as equivalent assuming hypothesized events non-occurred in reality to be ''counterfactual cases'' where a subset of these cases forms a contrastive explanation [10].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [10] source
Not checked in the free run.
- Premisecitation
Findings on explanation accumulated in humanities and social sciences show that explanation is intrinsically contrastive.
“Findings on explanation accumulated in humanities and social sciences show that it is intrinsically contrastive [11].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [11] source
Not checked in the free run.
- Premisecitation
Many existing XAI methods provide summaries of automatically made predictions rather than true explanations.
“In addition, many of the existing eXplainable AI (XAI 1 ) methods provide summaries of automatically made predictions rather than true explanations [4].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [4] source
Not checked in the free run.
- Premisecitation
Empirical studies prove that incorporating contrastiveness improves the quality of explanations offered to the end-user.
“Thus, empirical studies prove that incorporating contrastiveness improves the quality of explanations offered to the end-user [45].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [45] source
Not checked in the free run.
- Premisecitation
A lack of the ability to explain automatic decisions prevents users from a responsible exploitation of their decisions.
“Furthermore, it prevents users from a responsible exploitation of their decisions [3].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [3]
Not checked in the free run.
- Premisecitation
According to GDPR legal provisions, the data subject must be provided with meaningful information about the logic involved in automatic decision making, commonly referred to as the right to explanation.
“According to these legal provisions, the data subject must be provided with ''meaningful information about the logic involved'' in the automatic decision making process, which is commonly referred to as the ''right to explanation'' [6].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [6] source
Not checked in the free run.
- Premisecitation
Over et al. showed the existence of a connection between counterfactuals and conditional probability.
“In support of this assumption, Over et al. [40] showed the existence of connection between counterfactuals and conditional probability, as they experimented with probability judgments about counterfactuals.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [40]
Not checked in the free run.
- Premisecitation
Contrastive explanations can be used to personalize human-machine interaction when a user is engaged in an explanatory dialogue with an AI application.
“Furthermore, contrastive explanations can be used to personalize human-machine interaction when a user is engaged in an explanatory dialogue with an AI application.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [46] source
Not checked in the free run.
- Interpretationnone given
Counterfactual explanations are, in principle, model-agnostic, as they are suitable to explain the output of any black-box algorithm in a post-hoc manner.
“In addition, counterfactual explanations are, in principle, model-agnostic, as they are suitable to explain the output of any black-box algorithm in a post-hoc manner.”
Quote in the paperNot a substantial claim - Interpretationreasoning
An important advantage of counterfactual explanations over their non-counterfactual analogs is that they are devoid of any prerequisites to the data or model.
“An important advantage of counterfactual explanations over their non-counterfactual analogs is that they are devoid of any prerequisites to the data or model.”
Quote in the paperNot a substantial claim - Interpretationreasoning
Counterfactual explanations are data-agnostic as they can be based on the features of the neighbouring data examples extracted from the same training set and/or on the data generated synthetically around the data instance in question.
“Indeed, counterfactual explanations are dataagnostic as they can be based on the features of the neighbouring data examples extracted from the same training set and/or on the data generated synthetically around the data instance in question.”
Quote in the paperNot a substantial claim - Interpretationreasoning
Non-causal contrastive explanations can be used to explain the properties and relations inherent to such systems.
“Hence, they can be used to explain the properties and relations inherent to such systems.”
Quote in the paperNot a substantial claim - Premisecitation
The main focus of Russell et al.'s work [153] is placed on assessing coherence and diversity of generated counterfactuals.
“The main focus of the work is however placed on assessing coherence and diversity of generated counterfactuals.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [153] source
Not checked in the free run.
- Interpretationcitation
Counterfactual explanation generation appears highly relevant to sequential tasks in robotics such as automatic planning.
“Contfactual explanation generation appears highly relevant to sequential tasks in robotics such as automatic planning [94], [157], [161].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Premisecitation
Some of the robotics-related frameworks found in reinforcement learning settings provide explanations for policies that a robot selects at a given time step [132], [156].
“Moreover, some of the robotics-related frameworks found in reinforcement learning settings provide explanations for policies that a robot selects at a given time step [132], [156].”
Quote in the paperNot a substantial claim0 of 2 sources checked - Premisecitation
Guidotti et al. employ decision trees as part of reconstructing the reasoning behind any arbitrary classifier in a post-hoc fashion.
“In contrast to the model-specific frameworks operating on decision tree output, Guidotti et al. employ decision trees as part of reconstructing the reasoning behind any arbitrary classifier in a post-hoc fashion [140].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [140] source
Not checked in the free run.
- Premisecitation
Wachter et al. design a generic counterfactual explanation framework to find the closest point to the test data example.
“Wachter et al. design a generic counterfactual explanation framework to find the closest point to the test data example [6].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [6] source
Not checked in the free run.
- Premisecitation
Another case of counterfactual explanation generation regarded as an optimization problem is the ''Constrained Adversarial Examples'' framework.
“Another case of counterfactual explanation generation regarded as an optimization problem is the ''Constrained Adversarial Examples'' framework [148].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [148]
Not checked in the free run.
- Premisecitation
Dandl et al. formulate counterfactual search as a multiobjective optimization problem using a distance metric for mixed feature spaces aiming to obtain sparse and most plausible counterfactuals.
“Dandl et al. [134] formulate counterfactual search as a multiobjective optimization problem using a distance metric for mixed feature spaces aiming to obtain sparse and most plausible counterfactuals.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [134] source
Not checked in the free run.
- Premisecitation
The scalability of Anjomshoae et al.'s algorithm is limited to the use-cases operating on a small number of features.
“However, the scalability of such an algorithm is limited to the use-cases operating on a small number of features.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
- Premisecitation
van der Waa et al.'s Foil-Trees framework provides the methodological basis for perceptual-level contrastive explanation generation within the Perceptual-Cognitive Explanation framework.
“Furthermore, their ''Foil-Trees'' framework provides the methodological basis for perceptual-level contrastive explanation generation within the ''Perceptual-Cognitive Explanation'' framework [150].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [150] source
Not checked in the free run.
- Premisecitation
Numerical counterfactual explanations can be represented in tabular form reflecting necessary changes to affect the decision.
“or in tabular form reflecting necessary changes to affect the decision [55], [56], [61],”
Quote in the paperNot a substantial claim0 of 3 sources checked- cites [55]
Not checked in the free run.
- cites [56]
Not checked in the free run.
- cites [61] source
Not checked in the free run.
- Premisecitation
Visual counterfactual explanations for non-visual input data plot feature-value pair dependencies.
“On the one hand, visual explanations for non-visual input data (i.e., datasets containing continuous or categorical featurevalue pairs) plot feature-value pair dependencies [134], [141], [142].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Premisecitation
Numerical counterfactual explanations can be represented as logical formulas.
“They can be represented as logical formulas [67], [93], [94]”
Quote in the paperNot a substantial claim0 of 3 sources checked - Premisecitation
Multi-modal counterfactual explanations are often selected to be the most appropriate where the problem addressed is concerned with pairing a computer vision problem with a natural language processing task such as object detection and language grounding.
“They are often selected to be the most appropriate where the problem addressed is concerned with pairing a computer vision problem with a natural language processing task such as object detection and language grounding [86].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [86]
Not checked in the free run.
- Premisecitation
In Dhurandhar et al., the proximity of the selected counterexample to the test point is guaranteed by using an autoencoder.
“However, the proximity of the selected counterexample to the test point is, in this case, guaranteed by using an autoencoder.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [135] source
Not checked in the free run.
- Premisecitation
Mothilal et al. design a loss function with a diversity metric over the generated counterfactuals to provide end-users with multiple relevant counterfactual explanations.
“They design a loss function with a diversity metric over the generated counterfactuals to provide end-users with multiple relevant counterfactual explanations.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [105] source
Not checked in the free run.
- Premisecitation
Dhurandhar et al. address an optimization problem over a perturbation variable to produce a contrastive explanation for the image classification task.
“Similarly, Dhurandhar et al. address an optimization problem over a perturbation variable to produce a contrastive explanation for the image classification task [135].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [135] source
Not checked in the free run.
- Premisecitation
Mothilal et al. are also concerned with counterfactual diversity.
“Mothilal et al. are also concerned with counterfactual diversity [105].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [105] source
Not checked in the free run.
- Premisecitation
Numerical counterfactual explanations can replicate the internals of the classifier's structure, e.g., in the form of decision tree nodes or rules.
“In addition, they can replicate the internals of the classifier's structure, e.g., in the form of decision tree nodes or rules [80], [140], [152].”
Quote in the paperNot a substantial claim0 of 3 sources checked - Interpretationnone given
Certain application domains as well as the selection of a classifier require researchers to redefine counterfactuals imposing task-dependent constraints, which makes it nearly impossible to connect them to any of the existing theories of counterfactual explanation.
“However, certain application domains as well as the selection of a classifier require researchers to redefine contfactuals imposing task-dependent constraints, which makes it nearly impossible to connect them to any of the existing theories of contfactual explanation.”
Quote in the paperNot a substantial claim - Interpretationreasoning
In this interpretation, counterfactual explanations are conceptually most similar to counterfactuals as defined by Lewis.
“In this interpretation, counterfactual explanations are conceptually most similar to counterfactuals as defined by Lewis [36].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [36] source
Not checked in the free run.
- Classified as probably not a substantial claim. 205
- May combine two claims. 23
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenCausal relations between the variables in explanation are not always considered binary.
“Causal relations between the variables in explanation are not always considered binary (i.e., in the presence/absence of a cause).”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown data107 within-query duplicates were identified and removed from further analysis.
“107 within-query duplicates were identified and removed from further analysis.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenSeveral researchers attempt to unify causal and non-causal counterfactual explanation under the same paradigm.
“In addition, several researchers attempt to unify causal and non-causal contfactual explanation under the same paradigm.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown data33 studies were discarded after deeper analysis because they did not satisfy the inclusion criteria.
“enforced us to discard 33 studies which did not satisfy the inclusion criteria.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Findingnone givenCounterfactual explanations are used to justify automatic decisions obtained for a variety of AI-related problems.
“Contfactual explanations are used to justify automatic decisions obtained for a variety of AI-related problems.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
Findingown dataThe analysis of the primary studies related to RQ 2 allows categorization of the state-of-the-art counterfactual explanation generation frameworks according to four criteria: (1) the AI problem, (2) the explainability method, (3) the output representation, and (4) the evaluation method.
“The analysis of the primary studies related to RQ 2 allows us to categorize the state-of-the-art contfactual explanation generation frameworks in accordance with the following criteria: (1) the problem the solution for which is to be explained (i.e., the AI problem); (2) the method employed to generate such an explanation (i.e., the explainability method); (3) the output representation of the expl”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenIn three studies [67], [128], [137], the classifiers used in the experiments are not specified.
“In three studies [67], [128], [137], the classifiers used in the experiments are not specified.”
Quote in the paperNot a substantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown dataA stronger impact of counterfactual explanation in the results of the joint query q3 appears to affect significantly the overall allocation of related keywords in the corresponding keyword map.
“A stronger impact of counterfactual explanation in the results of the joint query q 3 appears to affect significantly the overall allocation of the related keywords in the corresponding keyword map (see Fig. 7).”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingown data75 publications were deemed irrelevant and filtered out during title, abstract, and author keyword screening.
“As shown in Fig. 2, 75 publications were deemed irrelevant and filtered out at this stage.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).
Findingown dataSimilar notions are essential for contrastive explanation.
“Similar notions are observed to be essential for contrastive explanation (see Fig. 6).”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
Probably not a substantial claim: may not pass S3 (load-bearing).
FindingreasoningA hybrid approach originating from monism has been adopted to unify causal and non-causal counterfactual explanation.
“Hence, a hybrid approach, originating from monism, 5 has been adopted to unify causal and non-causal counterfactual explanation.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenOther theoretical frameworks seek purely non-causal dependencies in explanation.
“In contrast, other theoretical frameworks seek purely non-causal dependencies in explanation.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenSeveral subcategorizations of causal counterfactuals have been suggested within some of the aforementioned theoretical frameworks.
“It is worth noting that several subcategorizations of causal contfactuals have been suggested within some of the aforementioned theoretical frameworks.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationIn the context of automatic decision-making, counterfactuals are most generally defined as counterarguments for an alternative prediction.
“In the context of automatic decision-making, counterfactuals are found to be most generally defined as counterarguments for an alternative prediction [86].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [86]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationOne of the classification-oriented frameworks [129] is extended to also handle the regression problem.
“One of the classificationoriented frameworks [129] is extended to also handle the regression problem, i.e., learning a mapping function f from a training dataset X to a continuous output variable Y .”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingown data19 papers were added to the review upon inspecting the bibliography of the primary studies.
“Finally, 19 papers were added to the review upon inspecting the bibliography of the primary studies.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingreasoningA distinction between different clusters in the contrastive explanation keyword map is visible more clearly.
“However, a distinction between different clusters in the latter case is visible more clearly.”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
Probably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationCausal contrastive explanation is frequently found to be designed as an answer to a why-question of the canonical form 'Why P rather than Q?', where P is an explanandum and Q is a foil.
“Causal contrastive explanation is frequently found to be designed as an answer to a why-question of the following canonical form: ''Why P rather than Q?'' where P is an explanandum (i.e., the fact to be explained), Q being a foil (i.e., one of alternative nonoccurring options) [21].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [21] source
Not checked in the free run.
- Rests on an assumption: Counterfactual explanation encompasses contrastive, counterfactual, and contrastive-counterfactual explanation.
Probably not a substantial claim: may not pass S3 (load-bearing).
FindingreasoningAll in all, 136 duplicates were removed.
“All in all, 136 duplicates were removed.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown data29 more duplicates were excluded when merging the sets of publications retrieved after q4 and q5.
“In addition, 29 more duplicates were excluded when merging the sets of publications retrieved after q 4 and q 5 .”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationSeveral explanation generators address computer vision tasks.
“For instance, several explanation generators address computer vision tasks.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
FindingcitationThe other frameworks addressing the regression problem aim to leverage gradient-boosted decision trees [61] and indicate how large errors in regression tasks could be overcome [56].
“The other frameworks addressing the regression problem aim to leverage gradientboosted decision trees [61] and indicate how large errors in regression tasks could be overcome [56].”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [61] source
Not checked in the free run.
- cites [56]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenAll the frameworks generating counterfactual explanations can be classified by their explainability method as either model-specific or model-agnostic.
“All the frameworks generating contfactual explanations can be classified by their explainability method as either modelspecific or model-agnostic.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenModel-agnostic frameworks are largely found to use decision trees as part of the reasoning mechanism instead of explaining their output.
“Model-agnostic frameworks are largely found to use decision trees as part of the reasoning mechanism instead of explaining their output.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenThe explanation generation is considered a two-fold optimization problem of finding pertinent positives and negatives.
“The explanation generation is considered a two-fold optimization problem of finding pertinent positives and negatives.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenCounterfactuals are presented in the form of intervals or specific values of the appropriate feature values whose alteration would have changed the output (numerical or feature-based output).
“Depending on the problem considered, contfactuals are presented in the form of: (1) intervals or specific values of the appropriate feature values whose alteration would have changed the output (i.e., numerical or featurebased output);”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenNumerical values (or intervals of values) associated to the most relevant features usually explain the behavior of AI algorithms.
“Numerical values (or intervals of values) associated to the most relevant features usually explain the behavior of AI algorithms.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenCounterfactuals are presented in the form of single- or multiple-sentence coherent text (linguistic output).
“(2) single-or multiple-sentence coherent text (i.e., linguistic output);”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown dataCounterfactuals are presented in the form of a multi-modal combination of some of the above.
“(4) a multi-modal combination of (some of) the above (see Table 8).”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenCounterfactuals are presented in the form of specific regions in the input image (visual output).
“(3) specific regions in the input image (i.e., visual output);”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationSome publications stress that human evaluation should be encouraged to estimate the quality and effectiveness of the generated counterfactuals.
“However, whereas certain publications do not provide any specific evaluation method, some do stress that human evaluation should be encouraged to estimate the quality and effectiveness of the generated counterfactuals [46], [150], [153].”
Quote in the paperNot a substantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenIn computer science and AI, notions of counterfactual and contrastive explanation are found to be the most dissimilar when applied to non-overlapping problems.
“In computer science and AI, notions of counterfactual and contrastive explanation are found to be the most dissimilar when applied to non-overlapping problems.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationThese changes are normally expected to be minimal.
“Furthermore, these changes are normally expected to be minimal [143].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [143] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenBoth types of evaluation methods highlight the features supporting evidence as formulated in the contrastive explanation on the basis of the values that a perturbation variable takes on.
“Both types of evaluation methods highlight the features supporting evidence as formulated in the contrastive explanation on the basis of the values that a perturbation variable takes on.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationTwo frameworks are evaluated in terms of both automatic metrics and human judgments.
“Two frameworks are evaluated in terms of both automatic metrics and human judgments.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
FindingreasoningDefinitions of counterfactual explanation vary depending on domains of science and even approaches used for solving specific tasks.
“In fact, definitions vary depending on domains of science and even approaches used for solving specific tasks.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenSubjective evaluation methods include human preferences for certain types of counterfactual explanation over others.
“The subjective methods include human preferences for certain types of contfactual explanation over others.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Findingnone givenContrastive and counterfactual explanations are found to be in great demand across various sub-fields of AI.
“Contrastive and counterfactual explanations are found to be in great demand across various sub-fields of AI.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationOnly linguistic templates for counterfactual explanations are defined without any theoretical grounding in accordance with any accounts described in Section IV-B.
“Alternatively, only linguistic templates for such explanations are defined without any theoretical grounding in accordance with any accounts described in Section IV-B.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Findingnone givenThe existing methodological differences affect greatly the definition of counterfactual explanation found across various approaches.
“In addition, the existing methodological differences affect greatly the definition of contfactual explanation found across various approaches.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Findingnone givenThe corresponding line of research in AI makes little use of the rich theoretical background accumulated by now.
“Nevertheless, the corresponding line of research in AI makes little use of the rich theoretical background accumulated by now.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenA consensus among researchers has been observed on how counterfactual explanations are defined irrespective of the theoretical framework proposed by individual authors.
“In general, a consensus among researchers has been observed on how contfactual explanations are defined irrespective of the theoretical framework proposed by individual authors.”
Quote in the paperNot a substantial claimMay combine two claims.
FindingreasoningThe concept of 'the closest possible world' does not always appear in the related publications, but it turns out to be implicitly wired in almost all works.
“Indeed, while the concept of ''the closest possible world'' does not always appear in the related publications, it turns out to be implicitly wired in almost all works.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingown dataThe topic of contrastive and counterfactual explanation has recently attracted attention from researchers in XAI.
“In addition, the topic has recently attracted attention from researchers in XAI (see Fig. 13).”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationThe most commonly acceptable definition of a counterfactual in the observed RQ3-related studies states that a counterfactual explanation is a set of minimal feature modifications that makes the model change the prediction.
“Nevertheless, the most commonly acceptable definition of a contfactual in the observed RQ 3 -related studies states that a contfactual explanation is a set of minimal feature modifications that makes the model change the prediction [81], [105], [122].”
- Also stated in: § VI. CONCLUDING REMARKS · plain
“the most commonly accepted definition of a contfactual explanation in the context of XAI refers to a minimal set of feature modifications that makes the model change the prediction.”
Quote in the paperNot a substantial claim0 of 3 sources checked- Also stated in: § VI. CONCLUDING REMARKS · plain
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenThe generated counterfactuals are inspected in terms of fairness which is calculated as the expected distance between the input and a counterfactual over distinct values for a specified feature set.
“In addition, the generated counterfactuals are inspected in terms of fairness which is calculated as the expected distance between the input and a counterfactual over distinct values for a specified feature set.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenIn image classification, it is found necessary to detect specific regions in the given test image.
“In image classification, it is found necessary to detect specific regions in the given test image.”
Quote in the paperNot a substantial claimMay combine two claims.
FindingcitationAs a majority of the considered frameworks are designed for tackling classification problems, counterfactual explanations operate on the notion of a contrast-class.
“As a majority of the considered frameworks are designed for tackling classification problems, contfactual explanations operate on the notion of a contrast-class (e.g., see [155]) answering the question: ''How is the prediction altered when the observation changes, given a classifier and an observation?''”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [155] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingnone givenThere is a number of marginal interpretations of counterfactuals among the RQ3-related studies.
“Finally, there is a number of marginal interpretations of contfactuals among the RQ 3 -related studies.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based).
Findingnone givenThe notion of counterfactual explanation is found to be highly task-and domain-specific.
“Despite the fact that the notion of contfactual explanation is found to be highly task-and domain-specific, the most commonly accepted definition of a contfactual explanation in the context of XAI refers to a minimal set of feature modifications that makes the model change the prediction.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Findingnone givenIn humanities and social sciences, a major difference in various counterfactual theories of explanation is observed to concern the causal nature of explanation and its extrapolation to non-causal cases.
“In humanities and social sciences, a major difference in various contfactual theories of explanation is observed to concern the causal nature of explanation and its extrapolation to non-causal cases.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationIt is worth noting that the run-time of explanation generation algorithms is reported in addition to the evaluation metrics for several frameworks.
“It is worth noting that the run-time of explanation generation algorithms is reported in addition to the evaluation metrics for several frameworks [132], [139], [146], [152], [156], [159].”
Quote in the paperNot a substantial claim0 of 3 sources checked- cites [132] source
Not checked in the free run.
- cites [139]
Not checked in the free run.
- cites [146], [152], [156], [159]
Not checked in the free run.
Classified as probably not a substantial claim.
Findingnone givenA gap between philosophical accounts of counterfactual explanation to scientific modeling and ML-related concepts makes the theoretical frameworks poorly applicable to XAI.
“We find that a gap between philosophical accounts of contfactual explanation to scientific modeling and ML-related concepts makes the theoretical frameworks poorly applicable to XAI.”
Quote in the paperSubstantial claimMay combine two claims.
FindingreasoningHendricks et al.'s human evaluation results only include evaluations for factual explanations and are therefore excluded from the taxonomy group being discussed.
“However, these only include evaluations for factual explanations and are therefore excluded from the taxonomy group being discussed.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingreasoningThe rest of works redefine counterfactual explanation on the basis of the problem-specific constraints without explicitly referring to the theoretical foundations described in Section IV-B.
“The rest of works redefine contfactual explanation on the basis of the problem-specific constraints without explicitly referring to the theoretical foundations described in Section IV-B.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationAutomatic metrics are complemented with a user study on intuitiveness, friendliness, comprehensibility, and understandability of generated explanations.
“Automatic metrics are complemented with a user study on intuitiveness, friendliness, comprehensibility, and understandability of generated explanations.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
FindingcitationML-specific counterfactual explanations are designed to answer the question: 'Why was the outcome Y observed instead of Y?'
“More precisely, ML-specific contfactual explanations are designed to answer the question: ''Why was the outcome Y observed instead of Y ?''[148].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [148]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationFang et al. referred to contrastive reasoning as a procedure operating on but-statements.
“Alternatively, Fang et al. [43] referred to contrastive reasoning as a procedure operating on ''butstatements'', as in ''all cars are polluting, but hybrid cars are not polluting'', which serves a principally different explanation generation task in comparison with the other aforementioned approaches.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [43] source
Not checked in the free run.
Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).
PremisecitationSupporters of the pragmatic approach to explanation argue that the ability to distinguish the answer to an explanatory question from a set of contrastive hypothesized alternatives provides the explainee with sufficiently comprehensive information on the reasoning behind the question.
“Thus, supporters of the pragmatic approach to explanation argue that it is exactly the ability to distinguish the answer to an explanatory question from a set of contrastive hypothesized alternatives that provides the explainee with sufficiently comprehensive information on the reasoning behind the question [13].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [13] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactuals describe events or states of the world that did not occur and implicitly or explicitly contradict factual world knowledge.
“Counterfactuals describe events or states of the world that did not occur and implicitly or explicitly contradict factual world knowledge [27].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [27] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationNo agreement about a definition of explanation has been reached despite the attention explanation has received in philosophy of science.
“However, no agreement about a definition of explanation has been reached despite the fact that explanation has called a significant amount of attention in, e.g., philosophy of science [7], [8].”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [7]
Not checked in the free run.
- cites [8]
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).
PremisecitationCounterfactual reasoning in computer science is defined as the process of evaluating conditional claims about alternative possibilities and their consequences.
“Thus, counterfactual reasoning in computer science is defined as the process of evaluating conditional claims about alternative possibilities and their consequences [33].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [33] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationContrastive reasoning forms the basis of abductive inference.
“In addition, contrastive reasoning forms the basis of abductive inference [19], i.e., the process of inferring certain facts that render some observation plausible [20].”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactual reasoning is claimed to be a key mechanism for explaining adaptive behavior in a changing environment.
“In addition, counterfactual reasoning is claimed to be a key mechanism for explaining adaptive behavior in a changing environment [25], [26].”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactuality is among the most fundamental concepts in theories of causation.
“Moreover, counterfactuality is among the most fundamental concepts in theories of causation [36], [37].”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [36] source
Not checked in the free run.
- cites [37]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe pragmatic approach is claimed to set a minimum criterion that an explanation must fulfill: it must favor the probability of the observed event P to all the hypothetical alternatives.
“This approach is also claimed to set a minimum criterion that an explanation must fulfill: it must favor the probability of the observed event P to all the hypothetical alternatives (Q 1 , Q 2 , . . ., Q n ) [14].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [14] source
Not checked in the free run.
Classified as probably not a substantial claim.
Study design and conductcitationThe present survey was undertaken as a systematic literature review following the guidelines by Kitchenham and Charters, Kitchenham et al., and Wohlin.
“The present survey has been undertaken as a systematic literature review following the guidelines by Kitchenham and Charters [58], Kitchenham et al. [59], and Wohlin [60].”
Quote in the paperSubstantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe property of contrastiveness presupposes that an explanation answers the why-question regarding the cause of the event in question in terms of hypothesized non-occurring alternatives.
“The property of contrastiveness presupposes that an explanation answers the given why-question regarding the cause of the event in question (''Why did P happen?'') in terms of hypothesized non-occurring alternatives (''Why did P happen rather than Q?'') [12].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [12] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe ability to explain a decision contrastively is claimed to lead to responsible decision-making.
“Finally, the ability to explain a decision contrastively is claimed to lead to responsible decision-making [47].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [47] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationOver et al.'s findings point to a strong correlation between the probability of the counterfactual conditional and causal strength judgments.
“Their findings point to a strong correlation between the probability of the counterfactual conditional and causal strength judgments.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [40]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationMcGill and Klein distinguished contrastive reasoning from its counterfactual counterpart.
“In contrast, McGill and Klein distinguished contrastive reasoning from its counterfactual counterpart [42].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [42]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe truth of a counterfactually inferred statement is resolved by modeling a situation where the smallest possible change in features of the actual world leads to a different state of things and estimating what is true in that setting.
“In this setting, the truth of a counterfactually inferred statement is resolved by: (1) modeling a situation where the smallest possible change in features of the actual world (as set in the antecedent) leads to a different (possibly, desired) state of things (the so-called ''closest'' or ''nearest'' possible world); and (2) estimating what is true in that setting [35].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [35] source
Not checked in the free run.
Classified as probably not a substantial claim.
Study design and conductnone givenThe review provides readers with a thorough and reproducible analysis of the interdisciplinary research field under study.
“In this work we conduct a systematic literature review which provides readers with a thorough and reproducible analysis of the interdisciplinary research field under study.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationPearl extended the definition of the causal counterfactual to estimate the probability of the truth of the consequent caused by the antecedent.
“Thus, Pearl extended the definition of the causal counterfactual to esti-mate the probability of the truth of the consequent caused by the antecedent (''a probability statement about the truth of y, had x been true, when it is known that y had been false when x was false'') [37].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [37]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationOver et al. proposed that the subjective probability of the counterfactual at the present time is the same as the conditional probability P(y|x) at some earlier time.
“Thus, they proposed that the subjective probability of the counterfactual at the present time is the same as the conditional probability P(y|x) at some earlier time.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [40]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe stochastic nature of predictions made by various AI algorithms is claimed to be among the main obstacles in reaching a true explanation.
“The stochastic nature of predictions made by various AI algorithms is claimed to be among the main obstacles in reaching a true explanation [44].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [44] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationEdgington regarded counterfactual judgments as uncertain conditional statements and therefore evaluated them by estimating their conditional probability given some endorsing event.
“On a similar note, Edgington regarded counterfactual judgments as uncertain conditional statements and therefore evaluated them by estimating their conditional probability given some endorsing event [41].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [41]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationContrastive explanations are claimed to be inherent to human cognition.
“Thus, contrastive explanations are claimed to be inherent to human cognition [16].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [16]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationA counterfactual definition of a cause of an arbitrary event traces back to Hume.
“A counterfactual definition of a cause of an arbitrary event traces back to Hume [38].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [38] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Premisenone givenCounterfactual explanations specify necessary minimal changes in the input so that a contrastive output is obtained.
“On the other hand, counterfactual explanations specify necessary minimal changes in the input so that a contrastive output is obtained.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactuals are argued to represent a causal relation between the event happened in reality and its imaginary counterpart.
“Indeed, counterfactuals are argued to represent a causal relation between the event happened in reality and its imaginary counterpart.”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [36] source
Not checked in the free run.
- cites [37]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactual conditional statements are claimed to enable people to produce utterances that are factually false yet truthful irrespective of the interpretation adopted.
“All in all, counterfactual conditional statements are claimed to enable people to produce utterances that are factually false yet truthful irrespective of the interpretation adopted [32].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [32] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn the context of AI, explanation often bases on judgments about why a certain outcome is predicted by an AI algorithm and hypotheses about causes with respect to given effects.
“In the context of AI, it often bases on judgments about why a certain outcome is predicted by an AI algorithm and hypotheses about causes with respect to given effects [10].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [10] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Premisenone givenThe need for explaining reasoning mechanisms behind AI applications is legally regulated in the European Union by the General Data Protection Regulation.
“For instance, the need for explaining reasoning mechanisms behind such applications is now legally regulated in the European Union by means of the General Data Protection Regulation.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Premisenone givenContrastive explanations point to the difference between the actual and a hypothetical decision.
“It is important to note that contrastive explanations point to the difference between the actual and a hypothetical decision.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe need to motivate automatic decisions with a clear explanation of why the algorithm outputs a particular decision has made the XAI research field grow quickly.
“As a result, the need to motivate automatic decisions with a clear explanation of why the algorithm outputs a particular decision has made the XAI research field grow quickly [5].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [5] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn its most general form, explanation is normally treated as an answer to the question of why something is the case.
“In its most general form, explanation is normally treated as ''an answer to the question of why something is the case'' [9].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [9]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationA counterfactual explanation of an ML-based algorithm prediction may describe the smallest change to the feature values that changes the prediction to a predefined output.
“For instance, a counterfactual explanation of an ML-based algorithm prediction may describe ''the smallest change to the feature values that changes the prediction to a predefined output'' [53].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [53] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe terms contrastive explanation and counterfactual explanation are sometimes used interchangeably in the context of XAI.
“However, these terms are sometimes used interchangeably in the context of XAI [48], [49].”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [48]
Not checked in the free run.
- cites [49] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataSix queries were performed over the title, abstract, and author keywords in Scopus and Web of Science.
“Subsequently, we performed six queries over the title, abstract, and author keywords in the aforementioned libraries (see the overall structure of the query pipeline in Fig. 1).”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductreasoningThe preprocessed collection of publications was split into two overlapping subsets to distinguish publications covering theoretical accounts of contfactual explanation and existing computational frameworks for contfactual explanation generation.
“Subsequently, the preprocessed collection of publications was split into two overlapping subsets aiming to distinguish the publications covering theoretical accounts of contfactual explanation with the aim of extracting the related definitions, theories (or their inferences or implications) (q 4 ) and existing computational frameworks for contfactual explanation generation (q 5 ).”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductreasoningA window span of three words was used to ensure that the attributes counterfactual and contrastive relate to explanation.
“In addition, we used a window span of three words (i.e., ''NEAR/3'') to ensure that the attributes ''counterfactual'' and ''contrastive'' relate to explanation.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationCounterfactual explanations are normally considered contrastive by nature and therefore present a source of valuable complementary information to a given automatic prediction.
“Counterfactual explanations are normally considered contrastive by nature and therefore present a source of valuable complementary information to a given automatic prediction [52].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [52] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Study design and conductnone givenPublications obtained as a result of q5 address RQ2.
“Similarly, publications obtained as a result of q 5 address RQ 2 .”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKeane et al. argue that applying case-based reasoning techniques for generating counterfactuals increases their explanatory competence.
“Furthermore, Keane et al. argue that applying case-based reasoning techniques for generating counterfactuals increases their explanatory competence [51].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [51]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenDuplicate reports of the same studies appearing in both Scopus and Web of Science were excluded.
“In contrast, we excluded duplicate reports of the same studies appeared in both Scopus and WoS.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductnone givenFor RQ2, publications not related to AI algorithms or applications and those where the proposed framework did not provide any human-comprehensible contfactual explanations as output were discarded.
“As for the publications related to RQ 2 , we discarded: (1) the publications which were not related to AI algorithms or applications as well as (2) those where the proposed framework did not provide any human-comprehensible contfactual explanations as output.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductnone givenFor RQ1, studies whose contents did not introduce any contfactual theory of explanation or containing no formal or informal definition of contrastive or counterfactual or contrastive-counterfactual explanation were removed.
“As for the publications related to RQ 1 , we also removed: (1) the studies whose contents did not introduce any contfactual theory of explanation or (2) those containing no formal or informal definition of contrastive or counterfactual or contrastivecounterfactual explanation.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim. May combine two claims.
Study design and conductreasoningFor RQ2, inclusion criteria required publications to present a novel approach, method, or framework for contfactual explanation generation whose output can serve to explain the reasoning of an AI algorithm and to be found in computer science, mathematics, or engineering.
“In order to capture existing computational frameworks for contfactual explanation generation, we included publications that: (1) present a novel approach, method, or framework for contfactual explanation generation whose output can serve to explain the reasoning of an AI algorithm and (2) are found in such subject areas as computer science, mathematics, engineering as well as in their sub-fields.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenThe search web tools of the selected digital libraries allow researchers to reproduce the original study.
“The search web tools of the selected digital libraries allow researchers to reproduce the original study.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductown dataThe search strategy selected the digital libraries Scopus and Web of Science to retrieve relevant publications.
“We selected the digital libraries Scopus and Web of Science (WoS) to retrieve relevant publications from.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn the context of XAI, an explanation for an automatic decision or prediction can be obtained abductively by attempting the search problem over the set of the known information concerning that observation.
“In the context of XAI, an explanation for an automatic decision or prediction, treated as an observation, can be obtained abductively by attempting the search problem over the set of the known information concerning that observation [50].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [50] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenThe number of publications in 2020 is limited to the search date.
“It is worth noting that the number of publications in 2020 is limited to the search date.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based). May combine two claims.
Study design and conductnone givenQueries q1 and q2 embrace all up-to-date publications containing mentions of counterfactual and contrastive explanation, respectively, found across all subject areas.
“Queries q 1 and q 2 embrace all the up-to-date publications containing mentions of counterfactual and contrastive explanation, respectively, found across all subject areas.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based). May combine two claims.
Premisenone givenScopus and Web of Science include research publications in computing and index studies across all scientific fields, which allows for an objective analysis of the interdisciplinary literature relevant to the research questions.
“These libraries do not only include research publications in computing but also index studies across all scientific fields, which allows for an objective analysis of the interdisciplinary literature relevant to the research questions posed.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductreasoningWord-stems were used to allow for maximal diversity of the retrieved papers.
“In order to capture all relevant publications, we only used the corresponding word-stems to allow for maximal diversity of the retrieved papers.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S2 (checkable).
InterpretationreasoningCounterfactuals are particularly suitable for informing the end-user why a given data example is assigned a particular class label.
“Indeed, contfactuals are particularly suitable for informing the end-user why a given data example is assigned a particular class label.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based).
Study design and conductnone givenPublications retrieved as a result of q4 form an exhaustive set of papers addressing RQ1.
“It is important to note that publications retrieved as a result of q 4 form an exhaustive set of papers addressing RQ 1 .”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactual explanation is widely addressed in the paradigm of case-based reasoning.
“Alternatively, counterfactual explanation is widely addressed in the paradigm of case-based reasoning, i.e., a family of problem solving methods based on appeals to precedent solutions.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [14] source
Not checked in the free run.
May combine two claims.
Interpretationown dataThe keywords identified in studies related to q3 testify that the issue of counterfactual explanation is highly interdisciplinary and finds application in both humanities and natural sciences.
“The keywords identified in the studies related to q 3 testify that the issue of contfactual explanation is highly interdisciplinary and finds application in both humanities and natural sciences.”
Quote in the paperNot a substantial claim- Rests on an assumption: Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductnone givenPapers that q6 returned address RQ3.
“Finally, the papers that q 6 returned address RQ 3 .”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Study design and conductnone givenThe use of the search web tools guarantees performing equivalent queries across both libraries.
“Furthermore, their use guarantees performing equivalent queries across both libraries.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Stepin et al.'s method [122], the generated counterfactuals are ranked using a XOR-based distance to find the most relevant counterfactual pertinent to the given contrast class.
“Subsequently, the generated counterfactuals are ranked using a XOR-based distance to find the most relevant counterfactual pertinent to the given contrast class.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [122] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationRussell et al. adopt a similar approach to encompass continuous and discrete variables as well as the combination of the two [153].
“Russell et al. adopt a similar approach to encompass continuous and discrete variables as well as the combination of the two [153].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [153] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Premisenone givenThe framework [93] is not designed to provide explanations for ML algorithms.
“It is worth noting that the framework is not designed to provide explanations for ML algorithms.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [93]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Hendricks et al.'s framework [86], the best counterfactual explanation is selected to be the most class-specific to the counterfactual image while being the most relevant to the input image.
“The best counterfactual explanation is then selected to be the most class-specific to the counterfactual image while being the most relevant to the input image.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [86]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationStepin et al.'s method [122] is further extended to generating counterfactuals for fuzzy decision trees.
“This method is further extended to generating counterfactuals for fuzzy decision trees.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [122] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationFernández et al. [80] present a recursive algorithm which extracts counterfactuals in the form of contrast-class decision tree nodes.
“For instance, Fernández et al. [80] present a recursive algorithm which extracts counterfactuals in the form of contrast-class decision tree nodes.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [80] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationHendricks et al. bind an input visual image with a paired textual counterfactual explanation generated by a recurrent neural network [86].
“Hendricks et al. bind an input visual image with a paired textual counterfactual explanation generated by a recurrent neural network [86].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [86]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationGhazimatin et al. obtain a counterfactual explanation by removing a minimal set of user actions so that the output recommendation changes.
“They obtain a counterfactual explanation by removing a minimal set of user actions so that the output recommendation changes.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [84] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn order to guarantee the coherence of the counterfactual data example used for explanation, Russell et al. propose an integer programming-based method [153].
“In order to guarantee the coherence of the counterfactual data example used for explanation, an integer programming-based method is proposed.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [153] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationRussell et al. claim that the generated counterfactual explanations are diverse, as diversity constraints are applied iteratively to a set of candidate counterfactuals [153].
“In addition, the generated counterfactual explanations are claimed to be diverse, as diversity constraints are applied iteratively to a set of candidate counterfactuals.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [153] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationMosca et al. introduce an argumentation-based framework for social network management [149].
“Mosca et al. introduce an argumentation-based framework for social network management [149].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [149] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Fernández et al.'s framework [80], the relevance of the generated counterfactuals is measured by calculating a variant of the Gower distance.
“The relevance of the generated counterfactuals is then measured by calculating a variant of the Gower distance.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [80] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationAguilar-Palacios et al. make use of the weighted Euclidean distance to present the forecast as a contrast to the neighbouring vectorized promotions.
“The researchers make use of the weighted Euclidean distance to present the forecast as a contrast to the neighbouring vectorized promotions.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [61] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe first of the considered frameworks (in chronological order) [93] offers explanations by reasoning abductively over the information extracted from a given knowledge base to answer a specific contrastive question.
“The first of the considered frameworks (in chronological order) [93] offers explanations by reasoning abductively over the information extracted from a given knowledge base to answer a specific contrastive question.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [93]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Akula et al.'s study [128], the explanations based on such graphs are passed on to human subjects to evaluate the contrastive answers to the predefined questions.
“The explanations based on such graphs are passed on to human subjects to evaluate the contrastive answers to the predefined questions.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [128] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationUstun et al. tackle the problem of actionability, i.e., constraining the generated counterfactuals in such a manner that the imposed changes ''do not alter immutable features'' and that they ''do not alter mutable features in an infeasible way'' [158].
“Thus, Ustun et al. tackle the problem of actionability, i.e., constraining the generated counterfactuals in such a manner that the imposed changes ''do not alter immutable features'' and that they ''do not alter mutable features in an infeasible way'' [158].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [158] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationIn Kanehira et al.'s framework [91], a counterfactual explanation is dependent on how likely a selected region in the given frame is classified positive and not negative, hence all such regions are scored and normalized.
“A counterfactual explanation is, in this case, dependent on how likely a selected region in the given frame is classified positive and not negative, hence all such regions are scored and normalized.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [91] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSokol and Flash's ''Glass-box'' framework, designed specifically for decision trees, is argued to be easily extendable to capture the output of other logical (rule-based) models.
“Designed specifically for decision trees, their ''Glass-box'' frame- work is argued to be easily extendable to capture the output of other logical (rule-based) models.”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn the framework [93], an explanation is considered to be a consistent set of disjunctive literals for the explanation-seeking question.
“In this setting, an explanation is considered to be a consistent set of disjunctive literals for the explanation-seeking question.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [93]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationGhazimatin et al. propose a graph-based recommendation system in a counterfactual setup [84].
“Ghazimatin et al. propose a graph-based recommendation system in a counterfactual setup [84].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [84] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Hendricks et al.'s framework [86], a number of candidate explanations (both image-relevant and non-relevant) are generated, paired, and ranked.
“In their framework, a number of candidate explanations (both image-relevant and non-relevant) are generated, paired, and ranked.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [86]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationTo approach the actionability problem, Ustun et al. employ a mixed integer programming method.
“To approach this problem, a mixed integer programming method is employed.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [158] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe proposed metric in Fernández et al.'s framework [80] penalizes the number of feature changes when traversing the tree so that sparsity is promoted.
“The proposed metric penalizes the number of feature changes when traversing the tree so that sparsity is promoted.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [80] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKanehira et al. train a post-hoc explanatory model to justify a video classifier's output [91].
“Alternatively, Kanehira et al. train a post-hoc explanatory model to justify a video classifier's output [91].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [91] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationAguilar-Palacios et al. generate contrastive explanations using gradient boosted decision trees to forecast promotional sales [61].
“Aguilar-Palacios et al. generate contrastive explanations using gradient boosted decision trees to forecast promotional sales [61].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [61] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSokol and Flash rely on the Manhattan distance measuring leaf-to-leaf distance in the tree to retrieve the most relevant counterfactuals [46], [155].
“Alternatively, Sokol and Flash rely on the Manhattan distance measuring leaf-to-leaf distance in the tree to retrieve the most relevant counterfactuals [46], [155].”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationMosca et al. use contrastive explanations to answer critical questions about agent actions in the context of multi-user privacy conflict.
“They use contrastive explanations to answer critical questions about agent actions in the context of multi-user privacy conflict.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [149] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationAkula et al. [128] present an empirical study where input video frames are paired with the corresponding AND-OR graphs, i.e., compositional recursively defined graph-based knowledge representations capturing contextual information.
“For instance, Akula et al. [128] present an empirical study where input video frames are paired with the corresponding AND-OR graphs, i.e., compositional recursively defined graphbased knowledge representations capturing contextual information.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [128] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationStepin et al. retrieve counterfactuals from a rule matrix where each rule is encoded in terms of all possible feature values [122].
“Stepin et al. retrieve counterfactuals from a rule matrix where each rule is encoded in terms of all possible feature values [122].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [122] source
Not checked in the free run.
May combine two claims.
PremisecitationRussell et al.'s framework [153] is limited to: (1) explaining predictions of only linear classifiers and (2) a simple structure of the textual explanation template.
“However, this framework is limited to: (1) explaining predictions of only linear classifiers and (2) a simple structure of the textual explanation template.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [153] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn the extended framework [129], the continuous output is subsequently converted to a lower-scale discrete value mapped to a textual description similar to that typical of a classification problem.
“However, the continuous output is, in this case, subsequently converted to a lower-scale discrete value mapped to a textual description similar to that typical of a classification problem.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Laugel et al., counterfactuals are selected from a local neighbourhood circling around the test example with the radius of the distance to the closest correctly predicted data point of a contrast-class.
“Counterfactuals are selected from a local neighbourhood circling around the test example with the radius of the distance to the closest correctly predicted data point of a contrast-class.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [144] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLabaien et al. generate contrastive explanations for time-series data.
“Labaien et al. generate contrastive explanations for time-series data [141].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [141] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationAnjomshoae et al. focus on contrastive explanations that maximize contextual importance and contextual utility.
“Anjomshoae et al. [129] focus on contrastive explanations that maximize contextual importance and contextual utility.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).
PremisecitationA similar limitation is observed due to possibly high variability of the input.
“A similar limitation is observed due to possibly high variability of the input.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationPlan explicability is reformulated in terms of its comprehensibility by an end-user.
“As plan explicability is reformulated in terms of its comprehensibility by an end-user.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [132] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSreedharan et al. consider the task of automatic analysis of counterfactual explanations in their ''Hierarchical Expertise-Level Modeling'' framework [156].
“Conversely, Sreedharan et al. consider the task of automatic analysis of counterfactual explanations in their ''Hierarchical Expertise-Level Modeling'' framework [156].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [156] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Kim et al.'s framework [94], the problem of obtaining contrastive explanations is designed as a Bayesian inference problem, with the posterior distribution to be maximized defined as the probability of a contrastive explanation given a set of positive and negative plan traces.
“The problem of obtaining contrastive explanations is designed as a Bayesian inference problem, with the posterior distribution to be maximized defined as the probability of a contrastive explanation given a set of positive and negative plan traces.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [94] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
Premisecitationvan der Waa et al. randomly sample or generate a data set in the neighbourhood local to the data point in question.
“Similarly, van der Waa et al. randomly sample or generate a data set in the neighbourhood local to the data point in question [159].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [159] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSukkerd et al.'s objective is to find an optimal policy that minimizes the expected cumulative cost of reaching a goal state over all closed policies.
“Hence, their objective is to find an optimal policy ''that minimizes the expected cumulative cost of reaching a goal state over all closed policies''.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [157] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe researchers attempt to find the best counterfactual explanation by minimizing the number of attributes changed.
“The researchers attempt to find the best counterfactual explanation by minimizing the number of attributes changed.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [148]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLaugel et al. argue that a synthesized counterfactual data point must be connected to the training data.
“They argue that a synthesized counterfactual data point must be connected to the training data.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [144] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationMartens and Provost suggest the model-agnostic Search for Explanations for Document Classification algorithm for retrieving counterfactual explanations.
“Hence, they suggest the model-agnostic ''Search for Explanations for Document Classification'' algorithm for retrieving counterfactual explanations.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [146] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationPawelczyk et al. make use of an autoencoder architecture for a pretrained classifier performing counterfactual search in the nearest neighbor style.
“Pawelczyk et al. make use of an autoencoder architecture for a pretrained classifier performing counterfactual search in the nearest neighbor style [151].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [151] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationZhao and Sukkerd explain an autonomous system's behaviour modeling it as a Markov decision process.
“In addition, Zhao and Sukkerd explain an autonomous system's behaviour modeling it as a Markov decision process [161].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [161]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationA decision tree is then trained to select the foil based on the minimum number of nodes between the original data point and the candidate foils.
“A decision tree is then trained to select the foil based on the minimum number of nodes between the original data point and the candidate foils.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [159] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationThe Search for Explanations for Document Classification algorithm is only directly applicable to binary linear classifiers, whereas heuristics are proposed for nonlinear models.
“However, it is only directly applicable to binary linear classifiers, whereas heuristics are proposed for nonlinear models.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [146] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Wachter et al., fixing the optimal set of weights of a trained classifier, the objective function minimizes the distance between the nearest data points of opposing classes.
“Fixing the optimal set of weights of a trained classifier, the objective function minimizes the distance between the nearest data points of opposing classes.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [6] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationMartens and Provost argue that decision trees are an inadequate tool for representing large documents.
“In contrast, Martens and Provost argue that decision trees are an inadequate tool for representing, e.g., large documents [146].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [146] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationIn Guidotti et al.'s Local Rule-based Explanation framework, they generate a local neighbourhood for the given pre-classified data example using a genetic algorithm and subsequently train a decision tree on that newly obtained dataset to select a minimally distant foil within that local neighbourhood.
“In their ''Local Rule-based Explanation'' framework, they generate a local neighbourhood for the given pre-classified data example using a genetic algorithm and subsequently train a decision tree on that newly obtained dataset to select a minimally distant foil within that local neighbourhood.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [140] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe explanation is believed to justify the rejection of the policies alternative to the optimal one.
“The explanation is believed to justify the rejection of the policies alternative to the optimal one.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [157] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe gradient direction is constrained to ensure the ethical adequacy of the explanation generated.
“Furthermore, the gradient direction is constrained to ensure the ethical adequacy of the explanation generated.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [148]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe context-based values are calculated for each feature used by a black-box model observing the changes in the output as the input varies across the range of all possible input values.
“Thus, the context-based values are calculated for each feature used by a black-box model observing the changes in the output as the input varies across the range of all possible input values.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationA contrastive explanation is presented as a product of the analysis of the optimal policy at the next time step and an opposing policy on the basis of the objective values.
“Thus, a contrastive explanation is presented as a product of the analysis of the optimal policy at the next time step and an opposing policy on the basis of the objective values.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [161]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe generated contrastive explanations are attributed to a specific group of users by means of ontology engineering at the cognitive level of the framework to make the explanations adaptive.
“Subsequently, the generated contrastive explanations are attributed to a specific group of users by means of ontology engineering at the cognitive level of the framework to make the explanations adaptive.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [150] source
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Interpretationnone givenIn accordance with the findings in the previous Section IV-C1, counterfactual explanations have a great potential for automatic planning-related tasks.
“In accordance with the findings in the previous Section IV-C1, contfactual explanations have a great potential for automatic planning-related tasks.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisereasoningAnjomshoae et al.'s framework is shown to be universally applicable to various classification and regression algorithms, being based on model-agnostic and problem-independent concepts.
“Being based on model-agnostic and problem-independent concepts, this framework is shown to be universally applicable to various classification and regression algorithms.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationSukkerd et al. formulate the planning problem as the shortest stochastic path problem and develop the corresponding problem solver to obtain a contrastive explanation.
“For instance, Sukkerd et al. formulate the planning problem as the shortest stochastic path problem and develop the corresponding problem solver to obtain a contrastive explanation [157].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [157] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationWachter et al. suggest the use of the Manhattan distance weighted by the inverse median absolute deviation to calculate the proximity of a counterfactual to the input data example.
“The researchers suggest the use of the Manhattan distance weighted by the inverse median absolute deviation to calculate the proximity of a counterfactual to the input data example.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [6] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKim et al. employ a Bayesian probabilistic model for generating contrastive explanations [94].
“For instance, Kim et al. employ a Bayesian probabilistic model for generating contrastive explanations [94].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [94] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLaugel et al. raise the issue of justification for counterfactual explanation.
“Laugel et al. raise the issue of justification for counterfactual explanation [144].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [144] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKim et al.'s framework [94] operates on a pair of plan traces defined in terms of linear temporal logic templates.
“Thus, the framework operates on a pair of plan traces defined in terms of linear temporal logic templates.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [94] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe robot's model is adapted to the updates of human's model of the problem.
“The robot's model is adapted to the updates of human's model of the problem.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [132] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn the Constrained Adversarial Examples framework, adversarial examples that could serve as the basis for the counterfactual explanation of the output of deep learning models are searched for with the aim of minimizing the loss with respect to the attributes (features) between the original and counterfactual data examples.
“Adversarial examples that could serve as the basis for the counterfactual explanation of the output of deep learning models are searched for with the aim of minimizing the loss with respect to the attributes (features) between the original and counterfactual data examples.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [148]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationChakraborti et al. formulate the multi-model planning problem as a tuple consisting of the planner's model of the problem and the corresponding human approximation thereof.
“In addition, Chakraborti et al. formulate the multi-model planning problem as a tuple consisting of the planner's model of the problem and the corresponding human approximation thereof [132].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [132] source
Not checked in the free run.
May combine two claims.
PremisecitationMultiple-sentence counterfactual explanations provide end-users with specific details for the given decision or explain multiple decisions at once.
“In contrast, multiple-sentence explanations provide end-users with specific details for the given decision [153], [156], [157], [161] or explain multiple decisions at once [146].”
Quote in the paperNot a substantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationFernández et al. introduce the notion of counterfactual sets to enhance counterfactual diversity.
“Fernández et al. introduce the notion of counterfactual sets to enhance counterfactual diversity [81].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [81] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual-linguistic counterfactual explanations allow for dialogic interaction.
“and allow for dialogic interaction [150].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [150] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual input data are explained by contrastive patterns between the given data example and that of an opposing class in one iteration or a series thereof.
“or explained by contrastive patterns between the given data example and that of an opposing class in one iteration [130], [144] or a series thereof [49]”
Quote in the paperNot a substantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSingle-sentence textual counterfactual explanations answer end-user's inquiries with respect to the automatic decision in question.
“They also answer end-user's inquiries with respect to the automatic decision in question [128], [159].”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationNumerical counterfactual explanations can be extracted from interpretable feature-value pairs as a result of pruning in the search space.
“They can be extracted from interpretable featurevalue pairs as a result of pruning in the search space [132].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [132] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe candidate counterfactuals are then clustered, as the initial local neighbourhood is updated to become a more extensive hyperspherical layer, until it can no longer be extended.
“The candidate counterfactuals are then clustered, as the initial local neighbourhood is updated to become a more extensive hyperspherical layer, until it can no longer be extended.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [144] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSingle-sentence textual counterfactual explanations suggest feature-value based instructions or alternative actions for a possible output change.
“Such explanations suggest featurevalue based instructions [48], [122], [131] or alternative actions for a possible output change [84], [149].”
Quote in the paperNot a substantial claim0 of 3 sources checkedProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationFernández et al. explain random forest predictions by fusing different tree predictors so that the resulting counterfactual set contains the most relevant counterfactual.
“They explain random forest predictions by fusing different tree predictors so that the resulting counterfactual set contains the most relevant counterfactual.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [81] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual-linguistic counterfactual explanations identify counterfactuality in videos.
“In addition, visual-linguistic explanations identify counterfactuality in videos [91]”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [91] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn Kusner et al., counterfactuals are presented in the form of conditional distributions and not structural equations despite the fact that the causal model employed follows Pearl's formalism.
“It is worth noting that counterfactuals are presented in the form of conditional distributions and not structural equations despite the fact that the causal model employed follows Pearl's formalism [37].”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [55]
Not checked in the free run.
- cites [37]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual input data are explained by depicting critical regions absent in the input data example that determine what lacks in the image to be classified differently.
“by depicting critical regions absent in the input data example that determine what lacks in the image to be classified differently [135]”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [135] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe other neighboring counterfactuals serve to diversify the output explanation.
“The other neighboring counterfactuals serve to diversify the output explanation.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [81] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationA textual explanation summarizes the degree of importance of the selected features, while a visual explanation may present contextual in-method metrics that justify the classifier's reasoning.
“While a textual explanation summarizes the degree of importance of the selected features, a visual explanation may present contextual in-method metrics that justify the classifier's reasoning [129].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [129] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationChang et al. employ a generative model to construct a saliency map while masking the other regions of the input image.
“They employ a generative model to construct a saliency map while masking the other regions of the input image.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [133] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn computer vision-related classification tasks, Chang et al. find the smallest region in the image whose substitution would change the classifier's prediction.
“In computer vision-related classification tasks, Chang et al. find the smallest region in the image whose substitution would change the classifier's prediction [133].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [133] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationCounterfactual explanations, as a mixture of tabular and visual output representations, appear also in an augmented reality framework.
“Contfactual explanations, as a mixture of tabular and visual output representations, appear also in an augmented reality framework [137].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [137] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSingle-sentence textual counterfactual explanations combine a textual description with explicitly stated numerical feature values.
“Single-sentence textual explanations combine a textual description with explicitly stated numerical feature values [6].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [6] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKusner et al. propose a causal model to assess the so-called ''counterfactual fairness''.
“Kusner et al. propose a causal model to assess the so-called ''counterfactual fairness'' [55].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [55]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual input data (images) are associated with saliency maps in counterfactual explanations.
“On the other hand, visual input data (i.e., images) are associated with saliency maps [133]”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [133] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLaugel et al. argue that the distance from the test instance to a counterfactual does not sufficiently measure counterfactual's relevance, as the counterfactual in question may appear disconnected from the ground-truth data.
“They argue that the distance from the test instance to a counterfactual does not sufficiently measure counterfactual's relevance, as the counterfactual in question may appear disconnected from the ground-truth data.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [100] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationLaugel et al. enhance the work on justified counterfactual explanations.
“Laugel et al. [100] enhance the work on justified counterfactual explanations.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [100] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationVisual input data are explained by visualizing spatial regions associated to data examples of opposed classes.
“or by visualizing spatial regions associated to data examples of opposed classes [139].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [139]
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIt is common to distinguish between intrinsic and extrinsic evaluation methods in natural language generation.
“For instance, it is common to distinguish between intrinsic and extrinsic evaluation methods in natural language generation [163].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [163] source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationGatt and Krahmer make a distinction between ''objective'' (automatic, corpus-based) and ''subjective'' (human judgements) metrics.
“In addition, Gatt and Krahmer make a distinction between ''objective'' (automatic, corpus-based) and ''subjective'' (human judgements) metrics [164].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [164] source
Not checked in the free run.
May combine two claims.
InterpretationreasoningSubjective metrics can be extrapolated to other (non-linguistic) modalities of generated explanations and are therefore used to form the basis of the evaluation method classification in this review.
“Thanks to their methodological universality, they can be extrapolated to other (non-linguistic) modalities of generated explanations (e.g., numerical, visual, or multi-modal) and are therefore used to form the basis of the evaluation method classification in this review.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
InterpretationreasoningEstablishing causal relations appears to be among the AI problems that are yet to attract global attention.
“As most of the current ML-tasks are centered around singling meaningful patterns out from unstructured data, establishing causal relations appears to be among the AI problems that are yet to attract global attention.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Interpretationnone givenExplanations of different modalities are not necessarily merged.
“Nevertheless, explanations of different modalities are not necessarily merged.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationInteraction with users can be enhanced by means of voice-based explanations combined with textual explanations.
“Finally, interaction with users can be enhanced by means of voice-based explanations combined with textual explanations [46], [155].”
Quote in the paperNot a substantial claim0 of 2 sources checkedProbably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Interpretationnone givenSeveral metrics can be extended to be applied to other approaches.
“Several metrics can be extended to be applied to other approaches.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
InterpretationcitationSome rule-based approaches used in expert systems are justified theoretically, but newly emerging tasks present novel challenges for theorists and call for updating the theories developed so far.
“While some rule-based approaches used in expert systems are justified theoretically (e.g., see [93]), newly emerging tasks present novel challenges for theorists and call for updating the theories developed so far.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [93]
Not checked in the free run.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.
Interpretationnone givenPromoting interactivity is expected to make explanation social and further increase user's trust in the system's output.
“Promoting interactivity (e.g., by engaging the end-user to participate in an explanatory dialogue with the system) is expected to make explanation social and further increase user's trust in the system's output.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim. May combine two claims.
Interpretationnone givenMainly applied to a wide range of tasks in computer vision and natural language processing, contrastive and counterfactual explanations present a powerful tool that enhances human-machine interaction and allows for further personalization of the output generated by various AI algorithms, including ML-based black-box algorithms.
“Mainly applied to a wide range of tasks in computer vision and natural language processing, they present a powerful tool that enhances human-machine interaction and allows for further personalization of the output generated by various AI algorithms, including ML-based black-box algorithms.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S2 (checkable), S4 (based).
Study design and conductreasoningThis research methodology allowed us to carry out an unbiased reproducible study from an interdisciplinary topic-specific search in reputable and trustworthy sources.
“This research methodology allowed us to carry out an unbiased reproducible study from an interdisciplinary topic-specific search in reputable and trustworthy sources.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
InterpretationreasoningThe amount of publications for RQ3 may be somewhat misleading, as counterfactual explanations are often redefined without specifically referring to theoretical contributors in explanation.
“However, the amount of publications for RQ 3 may be somewhat misleading, as contfactual explanations are often redefined without specifically referring to theoretical contributors in explanation.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe consistency of output explanations is shown by measuring their accuracy on the basis of mean error, mean absolute error, and mean absolute percentage error.
“The consistency of output explanations is otherwise shown by measuring their accuracy on the basis of mean error, mean absolute error, and mean absolute percentage error [61].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [61] source
Not checked in the free run.
The paper relies on this without support.
There is a gap between theoretical approaches to contrastive and counterfactual explanation and the corresponding computational frameworks.
“It is of crucial importance to bridge the gap between theoretical approaches to contrastive and counterfactual explanation and the corresponding computational frameworks.”
The paper relies on this without support.
Publication counts and author keyword maps are valid indicators of research attention and interdisciplinarity in the field.
“The bibliometric analysis over the queries q 1 , q 2 , and q 3 allows us to obtain a big picture of the research area of contfactual explanation generation and spot its key characteristics.”
“Author keyword maps allow us to present an overview of the terms most relevant to those specified in the preparatory queries (q 1 , q 2 , and q 3 ).”
The paper relies on this without support.
Counterfactual explanation encompasses contrastive, counterfactual, and contrastive-counterfactual explanation.
“Remind that contfactual explanation embraces contrastive, counterfactual, and contrastive-counterfactual explanation.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Explainability is a growing issue for financial institutions who seek to ensure that AI conclusions and procedures are easily comprehensible by bank officials.
“Furthermore, explainability is a growing issue for financial institutions who seek to ensure that AI conclusions and procedures are ''easily comprehensible'' by bank officials.”
From A Review of Trustworthy and Explainable Artificial Intelligence (XAI) (2023) · cites this paper as [51]
The Introduction says that the need to explain AI decisions clearly has driven rapid growth in explainable-AI research, and that AI applications should justify decisions comprehensively to end-users. It does not specifically discuss financial institutions or bank officials, so the claim's banking-specific framing is not supported. (LLM, conf=0.85)
Experimental — accuracy not yet measuredArticulating all of the system's components and operations is a barrier when trying to break the system down into its constituent aspects because all input variables have to be easy to comprehend.
“Since all of the input variables have to be easy to comprehend, articulating all of the system's components and operations is a barrier when trying to break the system down into its constituent aspects [45], [51].”
From A Review of Trustworthy and Explainable Artificial Intelligence (XAI) (2023) · cites this paper as [51]
The Introduction says that hard-to-interpret parameters make some AI outputs difficult to explain, and the Output Representation section discusses explanations using interpretable feature–value pairs. Neither passage says that all input variables must be easy to comprehend or that articulating every component and operation obstructs breaking a system into constituent aspects. Searches for those requirements and the proposed barrier found no supporting passage; page numbers were unavailable in our copy. (LLM, conf=0.74)
Experimental — accuracy not yet measuredCounterfactual explanations offer another avenue for achieving explanatory depth in XAI by providing insights into 'what-if' scenarios.
“In the work of Stepin et al. [141] counterfactual explanations offer another avenue for achieving explanatory depth in XAI by providing insights into”
From Recent Emerging Techniques in Explainable Artificial Intelligence to Enhance the Interpretable and Understanding of AI Models for Human (2025) · cites this paper as [141]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
All frameworks are said to be evaluated, but 17 are said not to be
MinorThe evaluation-method section says all considered frameworks are evaluated with intrinsic metrics, then states that 17 of the 52 publications do not evaluate their frameworks by either automatic metrics or a human survey.
“It is worth noting that all the considered frameworks are evaluated by means of intrinsic (either subjective or objective) metrics.”
“17 out of 52 (32.69%) of the considered publications do not evaluate their frameworks, i.e., neither automatic metrics for contfactual explanation generation are suggested nor a human evaluation survey is presented in such publications.”
Cites
Software survey: VOSviewer, a computer program for bibliometric mapping
van Eck, N. J., & Waltman, L. (2010) · Scientometrics · doi:10.1007/s11192-009-0146-3 · Cited by 21,484 · see citing works
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555 · see citing works
Rudin, C. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0048-x · Cited by 10,561 · see citing works
Barredo Arrieta, A., Díaz-Rodríguez, N., Del Ser, J., Bennetot, A., Tabik, S., Barbado, A., Garcia, S., Gil-Lopez, S., Molina, D., Benjamins, R., et al. (2020) · Information Fusion · doi:10.1016/j.inffus.2019.12.012 · Cited by 10,423 · cited 1× · see citing works
“and accountable (ensuring responsibility for the explanations provided) [57].”
BackgroundExplanation in artificial intelligence: Insights from the social sciences
Miller, T. (2019) · Artificial Intelligence · doi:10.1016/j.artint.2018.07.007 · Cited by 5,302 · see citing works
CAUSALITY: MODELS, REASONING AND INFERENCE
Hitchcock, C., & Pearl, J. (2001) · The Philosophical Review · doi:10.2307/3182612 · Cited by 4,824 · see citing works
Systematic Literature Reviews in Software Engineering—A Systematic Literature Review
Kitchenham, B., Pearl Brereton, O., Budgen, D., Turner, M., Bailey, J., & Linkman, S. (2009) · Information and Software Technology · doi:10.1016/j.infsof.2008.09.009 · Cited by 4,537 · see citing works
Wohlin, C. (2014) · Proceedings of the 18th International Conference on Evaluation and Assessment in Software Engineering · doi:10.1145/2601248.2601268 · Cited by 4,234 · cited 2× · see citing works
“The present survey has been undertaken as a systematic literature review following the guidelines by Kitchenham and Charters [58], Kitchenham et al. [59], and Wohlin [60].”
BackgroundMethodMaking things happen: a theory of causal explanation
(2004) · Choice Reviews Online · doi:10.5860/choice.41-5840 · Cited by 3,383 · see citing works
Causality: models, reasoning, and inference
(2000) · Choice Reviews Online · doi:10.5860/choice.38-1494a · Cited by 2,969 · see citing works
Bayesian inference for causal effects: The role of randomization
Rubin, D. B. (1978) · The Annals of Statistics · doi:10.1214/aos/1176344064 · Cited by 2,503 · see citing works
Stalnaker, R. C. (1968) · IFS · doi:10.1007/978-94-009-9117-0_2 · Cited by 2,220 · see citing works
Inference to the Best Explanation
Lipton, P. (2003) · Routledge eBooks · doi:10.4324/9780203470855 · Cited by 2,139 · see citing works
An Inquiry concerning human understanding
Hume, D. (1955) · Cited by 1,774 · see citing works
Explaining machine learning classifiers through diverse counterfactual explanations
Mothilal, R. K., Sharma, A., & Tan, C. (2020) · Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency · doi:10.1145/3351095.3372850 · Cited by 1,233 · see citing works
Roese, N. J. (1997) · Psychological Bulletin · doi:10.1037/0033-2909.121.1.133 · Cited by 1,156 · cited 2× · see citing works
“Cognitive scientists refer to such mental representations of alternatives to past events as counterfactuals (‘‘contrary-to-fact’’) [15].”
BackgroundCounterfactual explanations without opening the black box: automated decisions and the GDPR
Wachter, S., Mittelstadt, B., & Russell, C. (2017) · SSRN Electronic Journal · doi:10.2139/ssrn.3063289 · Cited by 1,061 · see citing works
Kusner, M. J., Loftus, J. R., Russell, C., & Silva, R. (2017) · arXiv · doi:10.48550/arxiv.1703.06856 · Cited by 995 · see citing works
An Enquiry Concerning Human Understanding
Hume, D. (1777) · Oxford University Press eBooks · doi:10.1093/oseo/instance.00046350 · Cited by 982 · see citing works
Gatt, A., & Krahmer, E. (2018) · Journal of Artificial Intelligence Research · doi:10.1613/jair.5477 · Cited by 776 · see citing works
Explanations based on the Missing: Towards Contrastive Explanations with Pertinent NegativesPreprint
Dhurandhar, A., Chen, P.-Y., Luss, R., Tu, C.-C., Ting, P.-S., Shanmugam, K., & Das, P. (2018) · Neural Information Processing Systems · doi:10.48550/arxiv.1802.07623 · Cited by 730 · see citing works
Backward induction and common knowledge of rationality
Aumann, R. J. (1995) · Games and Economic Behavior · doi:10.1016/s0899-8256(05)80015-6 · Cited by 588 · see citing works
Knowledge-based causal attribution: The abnormal conditions focus model.
Hilton, D. J., & Slugoski, B. R. (1986) · Psychological Review · doi:10.1037/0033-295x.93.1.75 · Cited by 503 · see citing works
Actionable Recourse in Linear Classification
Ustun, B., Spangher, A., & Liu, Y. (2019) · Proceedings of the Conference on Fairness, Accountability, and Transparency · doi:10.1145/3287560.3287566 · Cited by 501 · see citing works
Statistical Problems in Agricultural Experimentation
Neyman, J., Iwaszkiewicz, K., & Kołodziejczyk, S. (1935) · Journal of the Royal Statistical Society Series B (Statistical Methodology) · doi:10.2307/2983637 · Cited by 492 · see citing works
Byrne, R. M. J. (2015) · Annual Review of Psychology · doi:10.1146/annurev-psych-122414-033249 · Cited by 467 · see citing works
Conversational processes and causal explanation.
Hilton, D. J. (1990) · Psychological Bulletin · doi:10.1037/0033-2909.107.1.65 · Cited by 456 · see citing works
Explainable Agents and Robots: Results from a Systematic Literature Review
Anjomshoae, S., Najjar, A., Calvaresi, D., & Främling, K. (2019) · Adaptive Agents and Multi-Agent Systems · doi:10.65109/kczb5817 · Cited by 402 · see citing works
Inference to the Best Explanation
Lipton, P. (2010) · Taylor & Francis eBooks · doi:10.4324/9780203470855_chapter_4 · Cited by 390 · see citing works
Plan Explanations as Model Reconciliation: Moving Beyond Explanation as Soliloquy
Chakraborti, T., Sreedharan, S., Zhang, Y., & Kambhampati, S. (2017) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2017/23 · Cited by 359 · see citing works
Factual and Counterfactual Explanations for Black Box Decision Making
Guidotti, R., Monreale, A., Giannotti, F., Pedreschi, D., Ruggieri, S., & Turini, F. (2019) · IEEE Intelligent Systems · doi:10.1109/mis.2019.2957223 · Cited by 349 · see citing works
Lipton, P. (1990) · Royal Institute of Philosophy Supplement · doi:10.1017/s1358246100005130 · Cited by 343 · see citing works
How scientific models can explain
Bokulich, A. (2009) · Synthese · doi:10.1007/s11229-009-9565-1 · Cited by 332 · see citing works
Counterfactual Visual ExplanationsPreprint
Goyal, Y., Wu, Z., Ernst, J., Batra, D., Parikh, D., & Lee, S. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.07451 · Cited by 326 · see citing works
FACE: Feasible and Actionable Counterfactual Explanations
Poyiadzi, R., Sokol, K., Santos-Rodriguez, R., De Bie, T., & Flach, P. (2020) · Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3375627.3375850 · Cited by 317 · see citing works
Counterfactuals in Explainable Artificial Intelligence (XAI): Evidence from Human Reasoning
Byrne, R. M. J. (2019) · Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/876 · Cited by 285 · see citing works
Mental models and counterfactual thoughts about what might have been
Byrne, R. M. J. (2002) · Trends in Cognitive Sciences · doi:10.1016/s1364-6613(02)01974-5 · Cited by 282 · see citing works
Ylikoski, P., & Kuorikoski, J. (2008) · Philosophical Studies · doi:10.1007/s11098-008-9324-z · Cited by 280 · see citing works
The explanation game: Explaining machine learning models using shapley values
Merrick, L., & Taly, A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-57321-8_2 · Cited by 277 · see citing works
(1996) · Medical Entomology and Zoology · Cited by 276 · see citing works
Multi-Objective Counterfactual Explanations
Dandl, S., Molnar, C., Binder, M., & Bischl, B. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58112-1_31 · Cited by 271 · see citing works
Explaining Data-Driven Document Classifications1
Martens, D., & Provost, F. (2014) · MIS Quarterly · doi:10.25300/misq/2014/38.1.04 · Cited by 264 · see citing works
Human Digital Twin for Fitness Management
Barricelli, B. R., Casiraghi, E., Gliozzo, J., Petrini, A., & Valtolina, S. (2020) · IEEE Access · doi:10.1109/access.2020.2971576 · Cited by 259 · see citing works
Explanation and abductive inference
Lombrozo, T. (2012) · The Oxford Handbook of Thinking and Reasoning · doi:10.1093/oxfordhb/9780199734689.013.0014 · Cited by 242 · see citing works
Interventionist counterfactuals
Briggs, R. (2012) · Philosophical Studies · doi:10.1007/s11098-012-9908-5 · Cited by 241 · see citing works
Explanation in Case-Based Reasoning—Perspectives and Goals
Sørmo, F., Cassens, J., & Aamodt, A. (2005) · Artificial Intelligence Review · doi:10.1007/s10462-005-4607-7 · Cited by 239 · see citing works
The probability of causal conditionals
Over, D. E., Hadjichristidis, C., Evans, J. S. B. T., Handley, S. J., & Sloman, S. A. (2007) · Cognitive Psychology · doi:10.1016/j.cogpsych.2006.05.002 · Cited by 229 · see citing works
The Pragmatic Turn in Explainable Artificial Intelligence (XAI)
Páez, A. (2019) · Minds and Machines · doi:10.1007/s11023-019-09502-w · Cited by 229 · see citing works
Sharma, S., Henderson, J., & Ghosh, J. (2020) · Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3375627.3375812 · Cited by 226 · see citing works
Trust calibration within a human-robot team: Comparing automatically generated explanations
Wang, N., Pynadath, D. V., & Hill, S. G. (2016) · 2016 11th ACM/IEEE International Conference on Human-Robot Interaction (HRI) · doi:10.1109/hri.2016.7451741 · Cited by 220 · see citing works
Efficient Search for Diverse Coherent Explanations
Russell, C. (2019) · Proceedings of the Conference on Fairness, Accountability, and Transparency · doi:10.1145/3287560.3287569 · Cited by 218 · see citing works
Ginsberg, M. L. (1986) · Artificial Intelligence · doi:10.1016/0004-3702(86)90067-6 · Cited by 218 · see citing works
Anomalies in real and counterfactual worlds: An eye-movement investigation
Ferguson, H. J., & Sanford, A. J. (2008) · Journal of Memory and Language · doi:10.1016/j.jml.2007.06.007 · Cited by 198 · see citing works
The Dangers of Post-hoc Interpretability: Unjustified Counterfactual Explanations
Laugel, T., Lesot, M.-J., Marsala, C., Renard, X., & Detyniecki, M. (2019) · Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/388 · Cited by 194 · see citing works
Sokol, K., & Flach, P. (2020) · KI - Künstliche Intelligenz · doi:10.1007/s13218-020-00637-y · Cited by 186 · see citing works
(2018) · Oxford Scholarship Online · doi:10.1093/oso/9780198777946.001.0001 · Cited by 174 · see citing works
Hendricks, L. A., Hu, R., Darrell, T., & Akata, Z. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01216-8_17 · Cited by 171 · see citing works
Learning Model-Agnostic Counterfactual Explanations for Tabular Data
Pawelczyk, M., Broelemann, K., & Kasneci, G. (2020) · Proceedings of The Web Conference 2020 · doi:10.1145/3366423.3380087 · Cited by 167 · see citing works
The role of contrast in causal and explanatory claims
Hitchcock, C. (1996) · Synthese · doi:10.1007/bf00413843 · Cited by 166 · see citing works
Keane, M. T., & Smyth, B. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58342-2_11 · Cited by 158 · see citing works
Let Me Explain: Impact of Personal and Impersonal Explanations on Trust in Recommender Systems
Kunkel, J., Donkers, T., Michael, L., Barbu, C.-M., & Ziegler, J. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300717 · Cited by 156 · see citing works
What is the Source of Our Knowledge of Modal Truths?
Lowe, E. J. (2012) · Mind · doi:10.1093/mind/fzs122 · Cited by 150 · see citing works
Case Studies Nested in Fuzzy-set QCA on Sufficiency: Formalizing Case Selection and Causal Inference
Schneider, C. Q., & Rohlfing, I. (2016) · Sociological Methods & Research · doi:10.1177/0049124114532446 · Cited by 147 · see citing works
Cognitive Processes in Counterfactual Thinking about what Might Have Been
Byrne, R. M. J. (1997) · The Psychology of learning and motivation/The psychology of learning and motivation · doi:10.1016/s0079-7421(08)60501-0 · Cited by 146 · see citing works
Comparison-Based Inverse Classification for Interpretability in Machine Learning
Laugel, T., Lesot, M., Marsala, C., Renard, X., & Detyniecki, M. (2018) · Communications in computer and information science · doi:10.1007/978-3-319-91473-2_9 · Cited by 141 · see citing works
Explaining Image Classifiers by Counterfactual GenerationPreprint
Chang, C., Creager, E., Goldenberg, A., & Duvenaud, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1807.08024 · Cited by 130 · see citing works
Grahne, G. (1998) · Journal of Logic and Computation · doi:10.1093/logcom/8.1.87 · Cited by 123 · see citing works
Kulakova, E., & Nieuwland, M. S. (2016) · Language and Linguistics Compass · doi:10.1111/lnc3.12175 · Cited by 122 · see citing works
How Mathematics Can Make a Difference
Baron, S., Colyvan, M., & Ripley, D. (2017) · Research Bank (Australian Catholic University) · Cited by 120 · see citing works
Boutilier, C., & Beche, V. (1995) · Artificial Intelligence · doi:10.1016/0004-3702(94)00025-v · Cited by 118 · see citing works
van Fraassen, & Bas, C. (1980) · Cited by 117 · see citing works
Reutlinger, A. (2016) · Philosophy of Science · doi:10.1086/687859 · Cited by 116 · see citing works
Counterfactuals and Explanation
Kment, B. (2006) · Mind · doi:10.1093/mind/fzl261 · Cited by 111 · see citing works
Pitt, J. C. (1988) · Medical Entomology and Zoology · Cited by 107 · see citing works
Generative Counterfactual Introspection for Explainable Deep Learning
Liu, S., Kailkhura, B., Loveland, D., & Han, Y. (2019) · IEEE Global Conference on Signal and Information Processing · doi:10.1109/globalsip45357.2019.8969491 · Cited by 106 · see citing works
Abduction 101: Reasoning processes to aid discovery
Folger, R., & Stein, C. (2017) · Human Resource Management Review · doi:10.1016/j.hrmr.2016.08.007 · Cited by 101 · see citing works
HOW CONTRASTIVE EXPLANATION FACILITATES THEORY BUILDING.
Tsang, E. W. K., & Ellsaesser, F. (2011) · Academy of Management Review · doi:10.5465/amr.2009.0153 · Cited by 95 · see citing works
The Political Economy of Pork: Project Selection at The U.S. Army Corps of Engineers
Hird, J. (1991) · American Political Science Review · doi:10.2307/1963168 · Cited by 91 · see citing works
ViCE: visual counterfactual explanations for machine learning models
Gomez, O., Holter, S., Yuan, J., & Bertini, E. (2020) · Proceedings of the 25th International Conference on Intelligent User Interfaces · doi:10.1145/3377325.3377536 · Cited by 91 · see citing works
Perspectivism, inconsistent models, and contrastive explanation
Chakravartty, A. (2010) · Studies in History and Philosophy of Science Part A · doi:10.1016/j.shpsa.2010.10.007 · Cited by 88 · see citing works
Using Perceptual and Cognitive Explanations for Enhanced Human-Agent Team Performance
Neerincx, M. A., van der Waa, J., Kaptein, F., & van Diggelen, J. (2018) · Interacción · doi:10.1007/978-3-319-91122-9_18 · Cited by 88 · see citing works
PRINCE: Provider-side Interpretability with Counterfactual Explanations in Recommender Systems
Ghazimatin, A., Balalau, O., Saha Roy, R., & Weikum, G. (2020) · Proceedings of the 13th International Conference on Web Search and Data Mining · doi:10.1145/3336191.3371824 · Cited by 88 · see citing works
Machine learning: the new 'big thing' for competitive advantage
Attaran, M., & Deb, P. (2018) · International Journal of Knowledge Engineering and Data Mining · doi:10.1504/ijkedm.2018.095523 · Cited by 85 · see citing works
The Pragmatic Turn in Explainable Artificial Intelligence (XAI)
Páez, A. (2019) · PhilPapers (PhilPapers Foundation) · Cited by 80 · see citing works
Random forest explainability using counterfactual sets
Fernández, R. R., Martín de Diego, I., Aceña, V., Fernández-Isabel, A., & Moguerza, J. M. (2020) · Information Fusion · doi:10.1016/j.inffus.2020.07.001 · Cited by 73 · see citing works
Measurable Counterfactual Local Explanations for Any Classifier
Adam, W., & Artur, D. G. (2020) · Frontiers in artificial intelligence and applications · doi:10.3233/faia200387 · Cited by 73 · see citing works
External representations and scientific understanding
Kuorikoski, J., & Ylikoski, P. (2015) · Synthese · doi:10.1007/s11229-014-0591-2 · Cited by 71 · see citing works
Theory and Contrastive Explanation in Ethnography
Lichterman, P., & Reed, I. A. (2015) · Sociological Methods & Research · doi:10.1177/0049124114554458 · Cited by 71 · see citing works
Inference and Explanation in Counterfactual Reasoning
Rips, L. J., & Edwards, B. J. (2013) · Cognitive Science · doi:10.1111/cogs.12024 · Cited by 68 · see citing works
Processing counterfactual and hypothetical conditionals: An fMRI investigation
Kulakova, E., Aichhorn, M., Schurz, M., Kronbichler, M., & Perner, J. (2013) · NeuroImage · doi:10.1016/j.neuroimage.2013.01.060 · Cited by 66 · see citing works
Sokol, K., & Flach, P. A. (2018) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2018/865 · Cited by 65 · see citing works
Generalized Inverse Classification
Lash, M. T., Lin, Q., Street, W. N., Robinson, J. G., & Ohlmann, J. W. (2017) · Society for Industrial and Applied Mathematics eBooks · doi:10.1137/1.9781611974973.19 · Cited by 64 · see citing works
Le, T., Wang, S., & Lee, D. (2020) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3394486.3403066 · Cited by 57 · see citing works
Contrastive and counterfactual reasoning in causal judgment.
McGill, A. L., & Klein, J. G. (1993) · Journal of Personality and Social Psychology · doi:10.1037/0022-3514.64.6.897 · Cited by 56 · see citing works
Bayesian Inference of Linear Temporal Logic Specifications for Contrastive Explanations
Kim, J., Muise, C. J., Shah, A., Agarwal, S., & Shah, J. (2019) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/776 · Cited by 56 · see citing works
Sokol, K., & Flach, P. A. (2019) · Bristol Research (University of Bristol) · Cited by 53 · see citing works
Measurable Counterfactual Local Explanations for Any ClassifierPreprint
White, A., & Garcez, A. D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1908.03020 · Cited by 53 · see citing works
General theory of topological explanations and explanatory asymmetry
Kostić, D. (2020) · Philosophical Transactions of the Royal Society B · doi:10.1098/rstb.2019.0321 · Cited by 53 · see citing works
Hierarchical Expertise Level Modeling for User Specific Contrastive Explanations
Sreedharan, S., Srivastava, S., & Kambhampati, S. (2018) · Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2018/671 · Cited by 50 · see citing works
A Ranking‐Theoretic Approach to Conditionals
Spohn, W. (2013) · Cognitive Science · doi:10.1111/cogs.12057 · Cited by 49 · see citing works
A Counterfactual Approach to Explanation in Mathematics
Baron, S., Colyvan, M., & Ripley, D. (2020) · Philosophia Mathematica · doi:10.1093/philmat/nkz023 · Cited by 48 · see citing works
Explanations of Black-Box Model Predictions by Contextual Importance and Utility
Anjomshoae, S., Främling, K., & Najjar, A. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-30391-4_6 · Cited by 46 · see citing works
Lučić, A., Haned, H., & de Rijke, M. (2020) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3351095.3372824 · Cited by 46 · see citing works
Explanation and the theory of questions
Cross, C. B. (1991) · Erkenntnis · doi:10.1007/bf00385722 · Cited by 45 · see citing works
Does the counterfactual theory of explanation apply to non-causal explanations in metaphysics?
Reutlinger, A. (2017) · European Journal for Philosophy of Science · doi:10.1007/s13194-016-0155-z · Cited by 45 · see citing works
The Entropy Theory of Counterfactuals
Kutach, D. N. (2002) · Philosophy of Science · doi:10.1086/338942 · Cited by 44 · see citing works
Explaining Deep Learning Models with Constrained Adversarial Examples
Moore, J., Hammerla, N., & Watkins, C. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-29908-8_4 · Cited by 41 · see citing works
LoRMIkA: Local Rule-based Model Interpretability with k-optimal Associations
Rajapaksha, D. S., Bergmeir, C., & Buntine, W. L. (2019) · Information Sciences · doi:10.1016/j.ins.2020.05.126 · Cited by 40 · see citing works
Counterfactual Thinking as a Scientific Method
Wenzlhuemer, R. (2009) · Social Science Open Access Repository (GESIS – Leibniz Institute for the Social Sciences) · doi:10.12759/hsr.34.2009.2.27-56 · Cited by 38 · see citing works
Contrastive Explanations with Local Foil TreesPreprint
van der Waa, J., Robeer, M., van Diggelen, J., Brinkhuis, M. J. S., & Neerincx, M. A. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.07470 · Cited by 37 · see citing works
Social mechanisms and explanatory relevance
Ylikoski, P. (2011) · Cambridge University Press eBooks · doi:10.1017/cbo9780511921315.009 · Cited by 37 · see citing works
Stepin, I., Alonso, J. M., Catalá, A., & Pereira-Fariña, M. (2020) · IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) · doi:10.1109/fuzz48607.2020.9177629 · Cited by 35 · see citing works
Self-citationWhy P rather than Q? The curiosities of fact and foil
Barnes, E. (1994) · Philosophical Studies · doi:10.1007/bf00989743 · Cited by 33 · see citing works
Multimodal Explanations by Predicting Counterfactuality in Videos
Kanehira, A., Takemoto, K., Inayoshi, S., & Harada, T. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00879 · Cited by 33 · see citing works
Sokol, K., & Flach, P. A. (2020) · Explore Bristol Research · doi:10.48550/arxiv.2001.09734 · Cited by 33 · see citing works
Northcott, R. (2012) · Synthese · doi:10.1007/s11229-012-0124-9 · Cited by 32 · see citing works
Baron, S. (2019) · Mind · doi:10.1093/mind/fzz008 · Cited by 31 · see citing works
Situational Evidence: Strategies for Causal Reasoning From Observational Field Notes
Katz, J. (2015) · Sociological Methods & Research · doi:10.1177/0049124114554870 · Cited by 30 · see citing works
A non-technical explanation of the counterfactual definition of confounding.
Bours, M. (2020) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2020.01.021 · Cited by 30 · see citing works
Estimating Conditional Chances and Evaluating Counterfactuals
Edgington, D. (2013) · Studia Logica · doi:10.1007/s11225-013-9514-3 · Cited by 28 · see citing works
Unjustified Classification Regions and Counterfactual Explanations in Machine Learning
Laugel, T., Lesot, M.-J., Marsala, C., Renard, X., & Detyniecki, M. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-46147-8_3 · Cited by 28 · see citing works
Scientific explanation: A critical survey
Schurz, G. (1995) · Foundations of Science · doi:10.1007/bf00145406 · Cited by 27 · see citing works
Towards explainable multi-objective probabilistic planning
Sukkerd, R., Simmons, R., & Garlan, D. (2018) · Proceedings of the 4th International Workshop on Software Engineering for Smart Cyber-Physical Systems · doi:10.1145/3196478.3196488 · Cited by 27 · see citing works
The demand for contrastive explanations
Elzein, N. (2018) · Philosophical Studies · doi:10.1007/s11098-018-1065-z · Cited by 25 · see citing works
Explanatory relevance across disciplinary boundaries: the case of neuroeconomics
Kuorikoski, J., & Ylikoski, P. (2010) · Journal of Economic Methodology · doi:10.1080/13501781003756576 · Cited by 24 · see citing works
Mathematics and Explanatory Generality: Nothing but Cognitive Salience
Knowles, R., & Saatsi, J. T. (2019) · Erkenntnis · doi:10.1007/s10670-019-00146-x · Cited by 24 · see citing works
A Counterfactual Theory of Causal Explanation
Ruben, D.-H. (1994) · Noûs · doi:10.2307/2215475 · Cited by 23 · see citing works
Cold-Start Promotional Sales Forecasting Through Gradient Boosted-Based Contrastive Explanations
Aguilar-Palacios, C., Muñoz-Romero, S., & Rojo‐Álvarez, J. L. (2020) · IEEE Access · doi:10.1109/access.2020.3012032 · Cited by 23 · see citing works
An Inferential Account of Model Explanation
Fang, W. (2019) · Philosophia · doi:10.1007/s11406-018-9958-9 · Cited by 20 · see citing works
Zhang, Y., Paik, J., & Pirolli, P. (2015) · Topics in Cognitive Science · doi:10.1111/tops.12143 · Cited by 18 · see citing works
Contrastive Constraints Guide Explanation‐Based Category Learning
Chin‐Parker, S., & Cantelon, J. (2017) · Cognitive Science · doi:10.1111/cogs.12405 · Cited by 18 · see citing works
Machine Learning: The New 'Big Thing' for Competitive Advantage
Attaran, M., & Deb, P. (2018) · International Journal of Knowledge Engineering and Data Mining · doi:10.1504/ijkedm.2018.10015621 · Cited by 17 · see citing works
An ASP-Based Approach to Counterfactual Explanations for Classification
Bertossi, L. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-57977-7_5 · Cited by 17 · see citing works
Contrastive Explanations for a Deep Learning Model on Time-Series Data
Labaien, J., Zugasti, E., & De Carlos, X. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-59065-9_19 · Cited by 17 · see citing works
A contrastive account of explanation generation
Chin-Parker, S., & Bradner, A. (2017) · Psychonomic Bulletin & Review · doi:10.3758/s13423-017-1349-x · Cited by 15 · see citing works
Knowledge of objective modality
Strohminger, M., & Yli‐Vakkuri, J. (2018) · Philosophical Studies · doi:10.1007/s11098-018-1052-4 · Cited by 15 · see citing works
Against the Contrastive Account of Singular Causation
Steglich‐Petersen, A. (2011) · The British Journal for the Philosophy of Science · doi:10.1093/bjps/axr024 · Cited by 14 · see citing works
Contrastive Explanations to Classification Systems Using Sparse Dictionaries
Apicella, A., Isgrò, F., Prevete, R., & Tamburrini, G. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-30642-7_19 · Cited by 14 · see citing works
Natural Language Interaction with Explainable AI ModelsPreprint
Akula, A., Todorović, S., Chai, J. Y., & Zhu, S. C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1903.05720 · Cited by 13 · see citing works
Counterfactual explanations for data-driven decisions
Fernández, C., Provost, F. J., & Han, X. (2019) · Journal of the Association for Information Systems · Cited by 12 · see citing works
Organizational Closure Competencies and Scaling: A Realist Approach to Theorizing Social Enterprise
Seelos, C., & Mair, J. (2014) · Social Entrepreneurship and Research Methods · doi:10.1108/s1479-838720140000009014 · Cited by 11 · see citing works
Cognitive Prerequisites: The Special Case of Counterfactual Reasoning
Pereira, L. M., & Lopes, A. B. (2020) · Studies in applied philosophy, epistemology and rational ethics · doi:10.1007/978-3-030-39630-5_14 · Cited by 11 · see citing works
Agent EXPRI: Licence to Explain
Mosca, F., Sarkadi, Ş., Such, J. M., & McBurney, P. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-51924-7_2 · Cited by 11 · see citing works
Counterfactual Reasoning: A Basic Guide for Analysts, Strategists, and Decision Makers
Hendrickson, N. (2008) · PsycEXTRA Dataset · doi:10.1037/e509492010-001 · Cited by 11 · see citing works
ALEEDSA: Augmented Reality for Interactive Machine Learning
Ferrario, A., Weibel, R. P., & Feuerriegel, S. (2020) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3334480.3382937 · Cited by 10 · see citing works
Bergadano, F., Cutello, V., & Gunetti, D. (2000) · Springer eBooks · doi:10.1007/978-94-017-1733-5_5 · Cited by 9 · see citing works
Contrast, inference and scientific realism
Day, M., & Botterill, G. (2006) · Synthese · doi:10.1007/s11229-006-9117-x · Cited by 8 · see citing works
Agent-Causation, Explanation, and Akrasia: A Reply to Levy’s Hard Luck
Franklin, C. E. (2015) · Criminal Law and Philosophy · doi:10.1007/s11572-013-9274-1 · Cited by 7 · see citing works
A Quasi-Interventionist Theory Of Mathematical Explanation
Gijsbers, V. (2017) · Data Archiving and Networked Services (DANS) · doi:10.2143/lea.237.0.3197358 · Cited by 7 · see citing works
Interactive explanation for planning-based systems: WIP abstract
Zhao, E., & Sukkerd, R. (2019) · Proceedings of the 10th ACM/IEEE International Conference on Cyber-Physical Systems · doi:10.1145/3302509.3313322 · Cited by 7 · see citing works
Relevance Metric for Counterfactuals Selection in Decision Trees
Fernández, R. R., de Diego, I. M., Aceña, V., Moguerza, J. M., & Fernández-Isabel, A. (2019) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-33607-3_10 · Cited by 7 · see citing works
Causal Explanation and Explanatory Selection
Schweder, R. (1999) · Synthese · doi:10.1023/a:1005266706000 · Cited by 6 · see citing works
Explaining counterfactuals of freedom
Pruss, A. R., & Rasmussen, J. (2013) · Religious Studies · doi:10.1017/s0034412513000322 · Cited by 5 · see citing works
Counterfactual reasoning as a key for explaining adaptive behavior in a changing environment
Paik, J., Zhang, Y., & Pirolli, P. L. T. (2014) · Biologically Inspired Cognitive Architectures · doi:10.1016/j.bica.2014.11.004 · Cited by 5 · see citing works
REASONS, CAUSES, AND CONTRASTS
DICKENSON, J. (2007) · Pacific philosophical quarterly · doi:10.1111/j.1468-0114.2007.00278.x · Cited by 5 · see citing works
Scalar implicature and contrastive explanation
Chien, A. J. (2007) · Synthese · doi:10.1007/s11229-006-9153-6 · Cited by 5 · see citing works
A characterization of contrastive explanations computation
Kean, A. (1998) · Lecture notes in computer science · doi:10.1007/bfb0095304 · Cited by 4 · see citing works
Self-forming actions, contrastive explanations, and the structure of the will
Campbell, N. (2018) · Synthese · doi:10.1007/s11229-018-1749-0 · Cited by 4 · see citing works
Towards a Monist Theory of Explanation
Held, C. (2019) · Journal for General Philosophy of Science · doi:10.1007/s10838-019-09458-6 · Cited by 3 · see citing works
Manipulationism and causal exclusion
Pexton, M. (2017) · Philosophica · doi:10.21825/philosophica.82110 · Cited by 3 · see citing works
Right and Wrong Reasons in Folk‐Psychological Explanation
Botterill, G. (2009) · International Journal of Philosophical Studies · doi:10.1080/09672550903164442 · Cited by 2 · see citing works
Luck and Agent-Causation: A Response to Franklin
Levy, N. (2015) · Criminal Law and Philosophy · doi:10.1007/s11572-014-9306-5 · Cited by 2 · see citing works
Hypothetical Interventions and Belief Changes
Andreas, H., & Casini, L. (2019) · Foundations of Science · doi:10.1007/s10699-019-09605-8 · Cited by 2 · see citing works
Inference to the Best Explanation. Philosophical Issues in Science
Gibson, R. F. (1994) · The review of metaphysics · Cited by 2 · see citing works
A Method of Contrastive Reasoning with Inconsistent Ontologies
Fang, J., Huang, Z., & van Harmelen, F. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-29923-0_1 · Cited by 1 · see citing works
Causal Counterfactuals in Social Science Research
Holland, P. W. (2015) · Elsevier eBooks · doi:10.1016/b978-0-08-097086-8.44005-5 · Cited by 1 · see citing works
CONTRASTIVE EXPLANATIONS IN EVOLUTIONARY BIOLOGY
Boulter, S. (2012) · Ratio · doi:10.1111/j.1467-9329.2012.00555.x · Cited by 1 · cited 1× · see citing works
“Finally, Boulter illustrates the use of contrastive explanations to distinguish between actual and non-actual biological forms [70].”
BackgroundCapacities, explanation and the possibility of disunity
Hohwy, J. (2003) · International Studies in the Philosophy of Science · doi:10.1080/0269859031000160630 · Cited by 1 · see citing works
A method of contrastive reasoning with inconsistent ontologies
Fang, J., Huang, Z., & van Harmelen, F. (2012) · Cited by 1 · see citing works
Sober, E. (2020) · European Journal for Philosophy of Science · doi:10.1007/s13194-020-00299-5 · Cited by 1 · see citing works
Flagged first, then by citations
6 of the 176 references couldn’t be matched to a known paper.
Research works printed but not linked: 3
- lewis, 1973, Counterfactuals
- lewis, 1986, On the Plurality of Worlds
- molnar, 2019, Interpretable Machine Learning A Guide for Making Black Box Models Explainable
Cited by
Sorted by citations
Ali, S., Abuhmed, T., El-Sappagh, S., Muhammad, K., Alonso-Moral, J. M., Confalonieri, R., Guidotti, R., Del Ser, J., Díaz-Rodríguez, N., & Herrera, F. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.101805 · Cited by 1,762 · see citing works
Self-citationExplainable AI (XAI): A systematic meta-survey of current challenges and future opportunities
Saeed, W., & Omlin, C. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.110273 · Cited by 783 · see citing works
Survey of Explainable AI Techniques in Healthcare
Chaddad, A., Peng, J., Xu, J., & Bouridane, A. (2023) · Sensors · doi:10.3390/s23020634 · Cited by 591 · see citing works
Nauta, M., Trienes, J., Pathak, S., Nguyen, E., Peters, M., Schmitt, Y., Schlötterer, J., van Keulen, M., & Seifert, C. (2023) · ACM Computing Surveys · doi:10.1145/3583558 · Cited by 590 · see citing works
Counterfactual explanations and how to find them: literature review and benchmarking
Guidotti, R. (2022) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-022-00831-6 · Cited by 497 · see citing works
A Review of Trustworthy and Explainable Artificial Intelligence (XAI)
Chamola, V., Hassija, V., Sulthana, A. R., Ghosh, D., Dhingra, D., & Sikdar, B. (2023) · IEEE Access · doi:10.1109/access.2023.3294569 · Cited by 352 · see citing works
“Various client advocacy organizations, competitors, and internal stakeholders within banking firms all exhibit this upsurge [1], [51].”
BackgroundMethodSchwalbe, G., & Finzel, B. (2023) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-022-00867-8 · Cited by 381 · see citing works
Nazar, M., Alam, M. M., Yafi, E., & Su'ud, M. M. (2021) · IEEE Access · doi:10.1109/access.2021.3127881 · Cited by 312 · see citing works
Chou, Y.-L., Moreira, C., Bruza, P., Ouyang, C., & Jorge, J. (2022) · Information Fusion · doi:10.1016/j.inffus.2021.11.003 · Cited by 302 · see citing works
Kute, D. V., Pradhan, B., Shukla, N., & Alamri, A. (2021) · IEEE Access · doi:10.1109/access.2021.3086230 · Cited by 209 · see citing works
Pre-Trained Language Models for Text Generation: A Survey
Li, J., Tang, T., Zhao, W. X., Nie, J., & Wen, J.-R. (2024) · ACM Computing Surveys · doi:10.1145/3649449 · Cited by 304 · see citing works
Ding, W., Abdel‐Basset, M., Hawash, H., & Ali, A. M. A. (2022) · Information Sciences · doi:10.1016/j.ins.2022.10.013 · Cited by 196 · see citing works
Singhal, A., Neveditsin, N., Tanveer, H., & Mago, V. (2024) · JMIR Medical Informatics · doi:10.2196/50048 · Cited by 172 · see citing works
Explainable AI Over the Internet of Things (IoT): Overview, State-of-the-Art and Future Directions
Jagatheesaperumal, S. K., Pham, Q.-V., Ruby, R., Yang, Z., Xu, C., & Zhang, Z. (2022) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2022.3215676 · Cited by 119 · see citing works
A Survey of Algorithmic Recourse: Contrastive Explanations and Consequential Recommendations
Karimi, A.-H., Barthe, G., Schölkopf, B., & Valera, I. (2023) · ACM Computing Surveys · doi:10.1145/3527848 · Cited by 301 · see citing works
Mathew, D. E., Ebem, D. U., Ikegwu, A. C., Ukeoma, P. E., & Dibiaezue, N. F. (2025) · Neural Processing Letters · doi:10.1007/s11063-025-11732-2 · Cited by 168 · see citing works
Mehrotra, S., Degachi, C., Vereschak, O., Jonker, C. M., & Tielman, M. L. (2024) · ACM Journal on Responsible Computing · doi:10.1145/3696449 · Cited by 184 · see citing works
The coming of age of interpretable and explainable machine learning models
Lisboa, P. J. G., Saralajew, S., Vellido, A., Fernández-Domenech, R., & Villmann, T. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.02.040 · Cited by 155 · see citing works
Explainable AI for Safe and Trustworthy Autonomous Driving: A Systematic Review
Kuznietsov, A., Gyevnár, B., Wang, C., Peters, S., & Albrecht, S. V. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3474469 · Cited by 169 · see citing works
Model agnostic generation of counterfactual explanations for molecules
Wellawatte, G. P., Seshadri, A., & White, A. D. (2022) · Chemical Science · doi:10.1039/d1sc05259d · Cited by 143 · see citing works
EXplainable Artificial Intelligence (XAI)—From Theory to Methods and Applications
Ortigossa, E. S., Gonçalves, T., & Nonato, L. G. (2024) · IEEE Access · doi:10.1109/access.2024.3409843 · Cited by 124 · see citing works
On generating trustworthy counterfactual explanations
Del Ser, J., Barredo-Arrieta, A., Díaz-Rodríguez, N., Herrera, F., Saranti, A., & Holzinger, A. (2024) · Information Sciences · doi:10.1016/j.ins.2023.119898 · Cited by 122 · see citing works
Houssein, E. H., Gamal, A. M., Younis, E. M. G., & Mohamed, E. (2025) · Cluster Computing · doi:10.1007/s10586-025-05281-5 · Cited by 144 · see citing works
Mylrea, M., & Robinson, N. (2023) · Entropy · doi:10.3390/e25101429 · Cited by 103 · see citing works
Explainability and causability in digital pathology
Plass, M., Kargl, M., Kiehl, T., Regitnig, P., Geißler, C., Evans, T., Zerbe, N., Carvalho, R., Holzinger, A., & Müller, H. (2023) · The Journal of Pathology Clinical Research · doi:10.1002/cjp2.322 · Cited by 99 · see citing works
Cabitza, F., Campagner, A., Malgieri, G., Natali, C., Schneeberger, D., Stoeger, K., & Holzinger, A. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.118888 · Cited by 89 · see citing works
Explainable Machine Learning for Property Predictions in Compound Optimization
Rodríguez-Pérez, R., & Bajorath, J. (2021) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.1c01789 · Cited by 87 · see citing works
Contrastive Explanations for Model Interpretability
Jacovi, A., Swayamdipta, S., Ravfogel, S., Elazar, Y., Choi, Y., & Goldberg, Y. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.120 · Cited by 83 · see citing works
Nor, A. K. M., Pedapati, S. R., Muhammad, M. B., & Leiva, V. (2021) · Sensors · doi:10.3390/s21238020 · Cited by 79 · see citing works
Salvi, M., Seoni, S., Campagner, A., Gertych, A., Acharya, U. R., Molinari, F., & Cabitza, F. (2025) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2025.105846 · Cited by 84 · see citing works
Counterfactual Shapley Additive Explanations
Albini, E., Long, J., Dervovic, D., & Magazzeni, D. (2022) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3531146.3533168 · Cited by 74 · see citing works
Martins, T., de Almeida, A., Cardoso, E., & Nunes, L. (2023) · IEEE Access · doi:10.1109/access.2023.3347028 · Cited by 70 · see citing works
Niri, M. F., Aslansefat, K., Haghi, S., Hashemian, M., Daub, R., & Marco, J. (2023) · Energies · doi:10.3390/en16176360 · Cited by 60 · see citing works
The Intriguing Relation Between Counterfactual Explanations and Adversarial Examples
Freiesleben, T. (2021) · Minds and Machines · doi:10.1007/s11023-021-09580-9 · Cited by 58 · see citing works
Towards Data-And Knowledge-Driven AI: A Survey on Neuro-Symbolic Computing
Wang, W., Yang, Y., & Wu, F. Q. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3483273 · Cited by 92 · see citing works
Interpretable and explainable predictive machine learning models for data-driven protein engineering
Medina-Ortiz, D., Khalifeh, A., Anvari-Kazemabad, H., & Davari, M. D. (2024) · Biotechnology Advances · doi:10.1016/j.biotechadv.2024.108495 · Cited by 54 · see citing works
The Dawn of the Human-Machine Era: A forecast of new and emerging language technologies
Sayers, D., Sousa‐Silva, R., Höhn, S., Ahmedi, L., Allkivi-Metsoja, K., Anastasiou, D., Beňuš, Š., Bowker, L., Bytyçi, E., Catalá, A., et al. (2021) · doi:10.17011/jyx/reports/20210518/1 · Cited by 59 · see citing works
Counterfactual Explanations Can Be ManipulatedPreprint
Slack, D., Hilgard, S., Lakkaraju, H., & Singh, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.02666 · Cited by 55 · see citing works
Bridging the Transparency Gap: What Can Explainable AI Learn from the AI Act?
Gyevnár, B., Ferguson, N., & Schäfer, B. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230367 · Cited by 43 · see citing works
Alfeo, A. L., Zippo, A. G., Catrambone, V., Cimino, M. G. C. A., Toschi, N., & Valenza, G. (2023) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2023.107550 · Cited by 38 · see citing works
Wilson, A., Stefanik, C., & Shank, D. B. (2022) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2022.100229 · Cited by 38 · see citing works
Interpreting Language Models with Contrastive Explanations
Yin, K., & Neubig, G. (2022) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2022.emnlp-main.14 · Cited by 45 · see citing works
The Robustness of Counterfactual Explanations Over Time
Ferrario, A., & Loi, M. (2022) · IEEE Access · doi:10.1109/access.2022.3196917 · Cited by 38 · see citing works
R e LAX: Reinforcement Learning Agent Explainer for Arbitrary Predictive Models
Chen, Z., Silvestri, F., Wang, J., Zhu, H., Ahn, H., & Tolomei, G. (2022) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3511808.3557429 · Cited by 38 · see citing works
CARLA: A Python Library to Benchmark Algorithmic Recourse and Counterfactual Explanation AlgorithmsPreprint
Pawelczyk, M., Bielawski, S., van-den-Heuvel, J., Richter, T., & Kasneci, G. (2021) · NeurIPS Datasets and Benchmarks · doi:10.48550/arxiv.2108.00783 · Cited by 137 · see citing works
Nguyen, V. G., Sharma, P., Ağbulut, Ü., Le, H. S., Cao, D. N., Dzida, M., Osman, S. M., Le, H. C., & Trần, V. D. (2024) · International Journal of Green Energy · doi:10.1080/15435075.2024.2326076 · Cited by 36 · see citing works
Shang, R., Feng, K. J. K., & Shah, C. S. (2022) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3531146.3533189 · Cited by 37 · see citing works
Counterfactual Explanations for Models of Code
Cito, J., Dillig, I., Murali, V., & Chandra, S. (2021) · 2022 IEEE/ACM 44th International Conference on Software Engineering: Software Engineering in Practice (ICSE-SEIP) · doi:10.1145/3510457.3513081 · Cited by 66 · see citing works
Effects of explainable artificial intelligence in neurology decision support
Gombolay, G. Y., Silva, A., Schrum, M., Gopalan, N., Hallman‐Cooper, J., Dutt, M., & Gombolay, M. (2024) · Annals of Clinical and Translational Neurology · doi:10.1002/acn3.52036 · Cited by 34 · see citing works
Borrego-Díaz, J., & Galán-Páez, J. (2022) · Minds and Machines · doi:10.1007/s11023-022-09603-z · Cited by 34 · see citing works
On the robustness of sparse counterfactual explanations to adverse perturbations
Virgolin, M., & Fracaros, S. (2022) · Artificial Intelligence · doi:10.1016/j.artint.2022.103840 · Cited by 62 · see citing works
Zhang, Y., Song, S., Li, X., Gao, S., & Raubal, M. (2024) · Sustainable Cities and Society · doi:10.1016/j.scs.2024.105739 · Cited by 33 · see citing works
Gerlach, J., Hoppe, P. E., Jagels, S., Licker, L., & Breitner, M. H. (2022) · Electronic Markets · doi:10.1007/s12525-022-00603-6 · Cited by 33 · see citing works
Feature-Weighted Counterfactual-Based Explanation for Bankruptcy Prediction
Cho, S. H., & Shin, K.-S. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.119390 · Cited by 32 · see citing works
Explainable Machine Learning with Prior Knowledge: An Overview
Beckh, K., Müller, S., Jakobs, M., Toborek, V., Tan, H., Fischer, R., Welke, P., Houben, S., & von Rueden, L. (2021) · Lancaster EPrints (Lancaster University) · doi:10.48550/arxiv.2105.10172 · Cited by 40 · see citing works
FCE: Feedback Based Counterfactual Explanations for Explainable AI
Suffian, M., Graziani, P., Alonso, J. M., & Bogliolo, A. (2022) · IEEE Access · doi:10.1109/access.2022.3189432 · Cited by 32 · see citing works
Self-citationExplainable AI: To Reveal the Logic of Black-Box Models
Chinu, & Bansal, U. (2023) · New Generation Computing · doi:10.1007/s00354-022-00201-2 · Cited by 32 · see citing works
Mathematical optimization modelling for group counterfactual explanations
Carrizosa, E., Ramírez-Ayerbe, J., & Morales, D. R. (2024) · European Journal of Operational Research · doi:10.1016/j.ejor.2024.01.002 · Cited by 29 · see citing works
Pradeep, P., Caro-Martínez, M., & Wijekoon, A. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124733 · Cited by 28 · see citing works
A comprehensive review of explainable AI in cybersecurity: Decoding the black box
Sharma, A., Rani, S., & Shabaz, M. (2025) · ICT Express · doi:10.1016/j.icte.2025.10.004 · Cited by 28 · see citing works
Mariotti, E., Alonso, J. M., & Albert Gatt, A. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.101882 · Cited by 27 · see citing works
Self-citationProlog-based agnostic explanation module for structured pattern classification
N'apoles, G., Hoitsma, F., Knoben, A., Jastrzębska, A., & Leon, M. (2021) · Information Sciences · doi:10.1016/j.ins.2022.12.012 · Cited by 24 · see citing works
Chen, Z., Silvestri, F., Tolomei, G., Wang, J., Zhu, H., & Ahn, H. (2022) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2022.3223892 · Cited by 25 · see citing works
Gajcin, J., & Dusparić, I. (2024) · ACM Computing Surveys · doi:10.1145/3648472 · Cited by 24 · see citing works
Designing meaningful human oversight in AI
Zhu, L., Lu, Q., Ding, M., Lee, S. U., & Wang, C. (2026) · AI and Ethics · doi:10.1007/s43681-026-01147-7 · Cited by 25 · see citing works
Evaluating the effectiveness of XAI techniques for encoder-based language models
Mersha, M. A., Yigezu, M. G., & Kalita, J. K. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113042 · Cited by 23 · see citing works
Exploring the Role of Local and Global Explanations in Recommender Systems
Radensky, M., Downey, D., Lo, K., Popović, Z., & Weld, D. S. (2022) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3491101.3519795 · Cited by 27 · see citing works
Counterfactual explanations as interventions in latent space
Crupi, R., Castelnovo, A., Regoli, D., & González, B. S. M. (2022) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-022-00889-2 · Cited by 23 · see citing works
Counterfactual explanations for misclassified images: How human and machine explanations differ
Delaney, E. D., Pakrashi, A., Greene, D., & Keane, M. T. (2023) · Artificial Intelligence · doi:10.1016/j.artint.2023.103995 · Cited by 24 · see citing works
Benchmarking Instance-Centric Counterfactual Algorithms for XAI: From White Box to Black Box
Moreira, C., Chou, Y.-L., Hsieh, C., Ouyang, C., Pereira, J. M., & Jorge, J. A. (2024) · ACM Computing Surveys · doi:10.1145/3672553 · Cited by 22 · see citing works
Counterfactual Learning on Graphs: A Survey
Guo, Z., Wu, Z., Xiao, T., Aggarwal, C. C., Liu, H., & Wang, S. (2025) · Machine Intelligence Research · doi:10.1007/s11633-024-1519-z · Cited by 28 · see citing works
Factual and Counterfactual Explanations in Fuzzy Classification Trees
Fernandez, G., Aledo, J., Gamez, J., & Puerta, J. (2022) · IEEE transactions on fuzzy systems · doi:10.1109/tfuzz.2022.3179582 · Cited by 18 · see citing works
A Generalized Explanation Framework for Visualization of Deep Learning Model Predictions
Wang, P., & Vasconcelos, N. M. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3241106 · Cited by 21 · see citing works
Harnessing Prior Knowledge for Explainable Machine Learning: An Overview
Beckh, K., Müller, S., Jakobs, M., Toborek, V., Tan, H., Fischer, R., Welke, P., Houben, S., & von Rueden, L. (2023) · IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) · doi:10.1109/satml54575.2023.00038 · Cited by 22 · see citing works
Formalising the Robustness of Counterfactual Explanations for Neural Networks
Jiang, J., Leofante, F., Rago, A., & Toni, F. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i12.26740 · Cited by 22 · see citing works
Counterfactual Explanations for Machine Learning: Challenges RevisitedPreprint
Verma, S., Dickerson, J. P., & Hines, K. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.07756 · Cited by 30 · see citing works
SDA-Vis: A Visualization System for Student Dropout Analysis Based on Counterfactual Exploration
Garcia-Zanabria, G., Gutierrez-Pachas, D. A., Cámara-Chávez, G., Poco, J., & Gomez-Nieto, E. (2022) · Applied Sciences · doi:10.3390/app12125785 · Cited by 21 · see citing works
Learning Time Series Counterfactuals via Latent Space Representations
Wang, Z., Samsten, I., Mochaourab, R., & Papapetrou, P. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88942-5_29 · Cited by 21 · see citing works
Mapping the landscape of ethical considerations in explainable AI research
Nannini, L., Manerba, M. M., & Beretta, I. (2024) · Ethics and Information Technology · doi:10.1007/s10676-024-09773-7 · Cited by 16 · see citing works
Effective Explanations for Entity Resolution Models
Teofili, T., Firmani, D., Koudas, N., Martello, V., Merialdo, P., & Srivastava, D. (2022) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde53745.2022.00248 · Cited by 19 · see citing works
Glacier: guided locally constrained counterfactual explanations for time series classification
Wang, Z., Samsten, I., Miliou, I., Mochaourab, R., & Papapetrou, P. (2024) · Machine Learning · doi:10.1007/s10994-023-06502-x · Cited by 20 · see citing works
Explainable artificial intelligence for molecular design in pharmaceutical research
Lamens, A., & Bajorath, J. (2026) · Chemical Science · doi:10.1039/d5sc08461j · Cited by 19 · see citing works
On Explainable AI and Abductive Inference
Medianovskyi, K., & Pietarinen, A. (2022) · Philosophies · doi:10.3390/philosophies7020035 · Cited by 19 · see citing works
Stepin, I., Alonso-Moral, J., Catalá, A., & Pereira-Fariña, M. (2022) · Information Sciences · doi:10.1016/j.ins.2022.10.098 · Cited by 16 · see citing works
Self-citationEldrandaly, K. A., Abdel‐Basset, M., Ibrahim, M., & AbdelAziz, N. M. (2022) · Enterprise Information Systems · doi:10.1080/17517575.2022.2098537 · Cited by 17 · see citing works
Is the transposition of EU directives homogeneous? An AI counterfactual example
Ballestar, M. T., Sáinz, J., & Sanz, I. (2021) · Applied Economics Letters · doi:10.1080/13504851.2021.1967860 · Cited by 17 · see citing works
Zhang, M., Pan, Y., Cui, Q.-S., Lu, Y., & Yu, W.-H. (2025) · Experimental Gerontology · doi:10.1016/j.exger.2025.112812 · Cited by 19 · see citing works
Bauer, J. M., & Michalowski, M. (2025) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocaf110 · Cited by 17 · see citing works
An Overview of Explainable AI Methods, Forms and Frameworks
Kumar, D., & Mehta, M. A. (2022) · Intelligent systems reference library · doi:10.1007/978-3-031-12807-3_3 · Cited by 19 · see citing works
An interdisciplinary perspective on AI-supported decision making in medicine
Ammeling, J., Aubreville, M., Fritz, A., Kießig, A., Krügel, S., & Uhl, M. (2024) · Technology in Society · doi:10.1016/j.techsoc.2024.102791 · Cited by 17 · see citing works
From Explanation to Recommendation: Ethical Standards for Algorithmic Recourse
Sullivan, E., & Verreault-Julien, P. (2022) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3514094.3534185 · Cited by 16 · see citing works
Black-box error diagnosis in Deep Neural Networks for computer vision: a survey of tools
Fraternali, P., Milani, F., Torres, R. N., & Zangrando, N. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-08100-9 · Cited by 16 · see citing works
Confalonieri, R., & Alonso, J. M. (2023) · IEEE Intelligent Systems · doi:10.1109/mis.2023.3334639 · Cited by 16 · see citing works
Self-citationCounterfactual simulation in causal cognition
Gerstenberg, T. (2024) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2024.04.012 · Cited by 16 · see citing works
A Review of Interpretability Methods for Gradient Boosting Decision Trees
Gonçalves, V. S. F., & de Carvalho, V. R. (2025) · Journal of the Brazilian Computer Society · doi:10.5753/jbcs.2025.5324 · Cited by 15 · see citing works
Explaining black-box classifiers: Properties and functions
Leila Amgoud, A. (2023) · International Journal of Approximate Reasoning · doi:10.1016/j.ijar.2023.01.004 · Cited by 15 · see citing works
Oliveira, R., Sörensen, K., & Martens, D. (2023) · European Journal of Operational Research · doi:10.1016/j.ejor.2023.08.031 · Cited by 15 · see citing works
Diffeomorphic Counterfactuals With Generative Models
Dombrowski, A.-K., Gerken, J. E., Müller, K., & Kessel, P. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3339980 · Cited by 15 · see citing works
How to Build Self-Explaining Fuzzy Systems: From Interpretability to Explainability [AI-eXplained]
Stepin, I., Suffian, M., Catalá, A., & Alonso, J. M. (2024) · IEEE Computational Intelligence Magazine · doi:10.1109/mci.2023.3328098 · Cited by 15 · see citing works
Self-citationSuffian, M., Kuhl, U., Bogliolo, A., & Alonso-Moral, J. M. (2025) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2025.103484 · Cited by 14 · see citing works
- Authors
- Stepin, I., Alonso, J. M., Catala, A., & Pereira-Farina, M.
- Affiliations
- Universidade de Santiago de Compostela (ES)
- Venue
- IEEE Access
- Published
- 2021
- Volume
- 9
- Pages
- 11974-12001
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W3125997628
- Semantic Scholar
- e016ee7bfc73cb5b8a92f6c517389be837c035eb
- Cited by
- 376
- References
- 176
- Citations by year
- 2026: 47 · 2025: 80 · 2024: 75 · 2023: 79 · 2022: 55 · 2021: 39 · 2020: 1
- FWCI
- 21.91
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- IEEE Access (publishedVersion, cc-by-nc-nd)
- Repositorio institucional da Universidade de Santiago de Compostela (University of Santiago de Compostela) (submittedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Hispana (submittedVersion, cc-by-nc-nd)
- Funders
- Spanish Ministry of Science, Innovation and Universities (RTI2018-099646-B-I00, RED2018-102641-T)
- Galician Ministry of Education, University and Professional Training (ED431F 2018/02, ED431C 2018/29, ED431G/08, ED431G2019/04)
- European Regional Development Fund
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Keywords
- contrastive explanations, counterfactual explanations, explanation generation, explainable AI, evidence-based explanations, model interpretability
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-08.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
