Bias and Fairness in Large Language Models: A Survey
Open accessThe DOI matches its Crossref record; 313 of the 313 references it cites were found in a registry or an index; 151 of the 183 sources its claims cite were checked for support.
Abstract
Rapid advancements of large language models (LLMs) have enabled the processing, understanding, and generation of human-like text, with increasing integration into systems that touch our social sphere. Despite this success, these models can learn, perpetuate, and amplify harmful social biases. In this article, we present a comprehensive survey of bias evaluation and mitigation techniques for LLMs. We first consolidate, formalize, and expand notions of social bias and fairness in natural language processing, defining distinct facets of harm and introducing several desiderata to operationalize fairness for LLMs. We then unify the literature by proposing three intuitive taxonomies, two for bias evaluation, namely, metrics and datasets, and one for mitigation. Our first taxonomy of metrics for bias evaluation disambiguates the relationship between metrics and evaluation datasets, and organizes metrics by the different levels at which they operate in a model: embeddings, probabilities, and generated text. Our second taxonomy of datasets for bias evaluation categorizes datasets by their structure as counterfactual inputs or prompts, and identifies the targeted harms and social groups; we also release a consolidation of publicly available datasets for improved access. Our third taxonomy of techniques for bias mitigation classifies methods by their intervention during pre-processing, in-training, intra-processing, and post-processing, with granular subcategories that elucidate research trends. Finally, we identify open problems and challenges for future work. Synthesizing a wide range of recent research, we aim to provide a clear guide of the existing literature that empowers researchers and practitioners to better understand and prevent the propagation of bias in LLMs.
Overview
CoChat Summary
TL;DR
This survey consolidates, formalizes, and taxonomizes bias evaluation and mitigation for large language models, covering metrics, datasets, and mitigation techniques. It proposes three taxonomies and identifies open problems around power imbalances, fairness conceptualization, evaluation standards, mitigation scalability, and theoretical guarantees.
Key findings
- The survey proposes three taxonomies: bias evaluation metrics organized by what they use from a model (embeddings, probabilities, generated text); bias evaluation datasets organized by structure (counterfactual inputs vs. prompts); and bias mitigation techniques organized by intervention stage (pre-processing, in-training, intra-processing, post-processing).
- It formalizes social bias for LLMs as representational harms (e.g., derogatory language, disparate system performance, erasure, exclusionary norms, misrepresentation, stereotyping, toxicity) and allocational harms (direct and indirect discrimination), and introduces fairness desiderata including fairness through unawareness, invariance, equal social group associations, equal neutral associations, and replicated distributions.
- It reports limitations across metric classes: embedding-based and probability-based metrics may be only weakly or inconsistently correlated with downstream task bias, and generated-text metrics can be sensitive to decoding parameters; it also highlights reliability and validity issues in widely used datasets.
- It consolidates publicly available datasets, with reported sizes including CrowS-Pairs (1,508), StereoSet (16,995), BBQ (58,492), RealToxicityPrompts (100,000), BOLD (23,679), HolisticBias (460,000), and Bias NLI (5,712,066).
Methods
- Design: a comprehensive survey and taxonomy-building effort rather than an empirical study, synthesizing prior work on bias evaluation and mitigation for LLMs.
- Scope: focuses on bias issues in LLMs for English, with additional languages for machine translation and multilingual models, and restricts inclusion to works that propose novel closed-form metrics, datasets, or mitigation techniques and specify applicability to social bias or fairness.
- Analysis: organizes metrics, datasets, and mitigation techniques into taxonomies, formalizes several metrics and mitigation objectives with unified notation, discusses limitations of each class, and outlines open problems and recommendations.
Limitations
- The paper states that it aligns with a technical solutionist perspective rather than a critical theoretical one, and that its taxonomies are organized by technical implementation details instead of downstream usage contexts or harms.
- The paper states that it focuses narrowly on model training and evaluation, rather than a more holistic pipeline including problem formulation, data collection, and deployment/integration.
- The paper states that it is limited to English language papers; evident from the text, it also does not report a systematic search protocol, search dates, or screening counts.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures










Verification
Claims the paper makes
121 Valid · 12 Not valid · 18 Not checked · 106 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Despite the growing emphasis on addressing these issues, bias and fairness research in LLMs often fails to precisely describe the harms of model behaviors: who is harmed, why the behavior is harmful, and how the harm reflects and reinforces social principles or hierarchies (Blodgett et al. 2020).”
What the sources say
- Partly supportedcites Blodgett et al. 2020 matched to “Language (Technology) is Power: A Critical Survey of “Bias” in NLP” (2020) sourceExperimental — accuracy not yet measured
Section 3.1 (Motivations; page not provided) finds that papers on bias in NLP often fail to explain which model behaviors are harmful, how, to whom, and why. Sections 4.1–4.2 call for attention to social hierarchies and explicit normative reasoning. The survey concerns NLP research broadly, however, and does not establish that this failure is common specifically in LLM research. (LLM, conf=0.86)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Similar to the shortcomings of embedding-based metrics, Delobelle et al. (2022) and Kaneko, Bollegala, and Okazaki (2022) point out that probability-based metrics may be only weakly correlated with biases that appear in downstream tasks”
- Also stated in: § 3.6 Recommendations · p. 27 · plain
“Probability-based metrics also show weak correlations with downstream biases.”
What the sources say
- Supportedcites Delobelle et al. 2022 matched to “Measuring Fairness with Biased Rulers: A Comparative Study on Bias Metrics for Pre-trained Language Models” (2022) source
Section 3 describes CrowS-Pairs as a pseudo-loglikelihood-based metric. Section 4.4 reports that CrowS-Pairs correlates less with two measures of bias in downstream tasks, supporting the hedged claim that probability-based metrics may correlate weakly with downstream bias. The finding does not establish that all probability-based metrics do so. (LLM, conf=0.90)
- Supportedcites Kaneko, Bollegala, and Okazaki 2022 matched to “Debiasing isn't enough! -- On the Effectiveness of Debiasing MLMs and their Social Biases in Downstream Tasks” (2022) source
The Introduction (p. 1) identifies intrinsic measures as using likelihood scores assigned by MLMs and reports only weak correlations with downstream, task-specific bias measures. Section 5.2 (p. 8) likewise finds weak and inconsistent correlations between intrinsic and extrinsic bias scores. (LLM, conf=0.93)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 3.6 Recommendations · p. 27 · plain
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Many approaches, for instance, assume some implicitly desirable criterion ( e.g. , a model output should be independent of any social group in the input), but do not explicitly acknowledge or state the normative social values that justify their framework.”
What the sources say
- Partly supportedcites Blodgett et al. 2021 matched to “Stereotyping Norwegian Salmon: An Inventory of Pitfalls in Fairness Benchmark Datasets” (2021) sourceExperimental — accuracy not yet measured
Section 2.3 describes assumptions about ideal model behavior, including equal probabilities for sentences about different target groups. Section 6 says the benchmarks make implicit decisions about stereotyping that are not explicitly discussed or justified. The paper examines four stereotyping benchmarks, however, rather than establishing this claim about many bias and fairness approaches generally or specifically about their normative social values. (LLM, conf=0.84)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Cabello, Jørgensen, and Søgaard (2023) illustrate that associations between the representations of protected attribute and other words can be independent of downstream performance disparities, if certain assumptions of social groups' language use are violated.”
What the sources say
- Supportedcites Cabello, Jørgensen, and Søgaard 2023 matched to “On the independence of association bias and empirical fairness in language models” (2023) source
The Bias section defines association bias as differences in how demographic-group terms are encoded and empirical fairness as performance differences across social groups. Section 5 says the two can be independent and explains that linking them relies on the In-Group Affinity Assumption about how groups use those terms; it calls that assumption problematic. No page numbers are provided in our copy. (LLM, conf=0.93)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Each stakeholder brings perspectives that may specify different fairness constraints for the same application and setting.”
What the sources say
- Partly supportedcites Jernite et al. (2022) matched to “Data Governance in the Age of Large-Scale Data-Driven Language Technology” (2022) sourceExperimental — accuracy not yet measured
Section 2.2 (p. 4) says participants bring different perspectives and warns that relying only on their values could exclude views more relevant to other language users. The paper discusses nondiscrimination and local expressions of values, but does not say that stakeholders specify different fairness constraints for the same application and setting. (LLM, conf=0.85)
- Not matchedcites Bender (2019)
An author–year citation the paper’s reference list doesn’t settle. We don’t search these by name.
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
71 more claims · 34 with their cited sources checked
Create a free account to see all 257 claims and every source verdict.
- Probably not a substantial claim. 151
- It may state two separate claims as one. 70
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“While this process mitigates some risks inherent to AI-generated content, it does not eliminate potential biases associated with proprietary large language models, which lack transparency regarding training data and internal parameters [10].”
From AI-Generated Multiple Mini Interview (MMI) Stations for Medical School Admissions: Psychometric Evaluation (2026) · cites this paper as [10]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Notably, there is growing attention on assessing LLM bias [63-64] and robustness [323-324].”
From A Survey of Large Language Models (2026) · cites this paper as [63]
The cited source says what the claim attributes to it.
In the citing paper
“these models pose serious risks, as the vast, unfiltered internet data used to train them can reinforce and amplify harmful biases, disproportionately harming marginalized communities (73).”
From Emergent social conventions and collective bias in LLM populations (2025) · cites this paper as [73]
The cited source says what the claim attributes to it.
In the citing paper
“Despite the use of various mitigation techniques, such as debiasing algorithms, curating balanced datasets, or incorporating fairness-focused training objectives, eliminating bias entirely is a persistent challenge141–143.”
From Current applications and challenges in large language models for patient care: a systematic review (2025) · cites this paper as [142]
The cited source says what the claim attributes to it.
In the citing paper
“Another major concern with LLMs is their propensity to amplify historical biases.”
From Evaluation of Large Language Models for Peer Review in Transplantation Research: Algorithm Validation Study (2026) · cites this paper as [30]
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
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.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 4.1 Counterfactual Inputs · p. 28
“| WinoBias+ | 1,367”
§ 4.1 Counterfactual Inputs · p. 29
“WinoBias+ (Vanmassenhove, Emmery, and Shterionov 2021) augments WinoBias with gender-neutral alternatives, similar to Winogender's neutral option, with 3,167 total instances.”
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 6×
“, 2020), and GPT-4 (OpenAI, 2023); autoencoding models include BERT (Devlin et al., 2019), RoBERTa (Liu et al.”
BackgroundInfluential(2014) · Cited by 33,419
A Unified Approach to Interpreting Model Predictions
Lundberg, S., & Lee, S.-I. (2017) · arXiv · doi:10.48550/arxiv.1705.07874 · Cited by 23,707
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 2×
“, 2019), GPT-3 (Brown et al., 2020), and GPT-4 (OpenAI, 2023); autoencoding models include BERT (Devlin et al.”
Background"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
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515
Kaugeranna, R., Kaugeranna, E., & Aion, (. S. 4. 6. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.giscience.2023.43 · Cited by 13,690
Gauriat, C.-M., Pencolé, Y., Ribot, P., & Brouillet, G. (2024) · HAL (Le Centre pour la Communication Scientifique Directe) · doi:10.4230/oasics.dx.2024.27 · Cited by 12,586
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929 · cited 1×
“The rise and rapid advancement of large language models (LLMs) has fundamentally changed language technologies (e.g., Brown et al., 2020; Conneau et al., 2020; Devlin et al., 2019; Lewis et al., 2020; Liu et al., 2019; OpenAI, 2023; Radford et al., 2018; 2019; Raffel et al., 2020).”
BackgroundMeasuring individual differences in implicit cognition: the implicit association test
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998) · Journal of Personality and Social Psychology · doi:10.1037/0022-3514.74.6.1464 · Cited by 9,210 · cited 2×
“, family and occupation words), emulating the Implicit Association Test (Greenwald et al., 1998).”
Method
Research works printed but not linked: 5
- Barocas, 2019, Fairness and Machine Learning: Limitations and Opportunities
- Blodgett, Su Lin . 2021. Sociolinguistically Driven Approaches for Just Natural Language Processing. Ph.D. thesis. University of Massachusetts Amherst.
- Green, 2019, ”Good” isn’t good enough, 1
- Radford, 2019, Language models are unsupervised multitask learners, OpenAI Blog, 1, 9
- Wang, 2019, BERT has a mouth, and it must speak: BERT as a Markov random field language model, 30, 10.18653/v1/w19-2304 · doi:10.18653/v1/w19-2304
Cited by
Items 301–400 of 763
LLM-based Multi-Agent Collaboration for Abstract Screening towards Automated Systematic ReviewsPreprint
Akinseloyin, O., Jiang, X., & Palade, V. (2025) · medRxiv · doi:10.1101/2025.08.11.25333429 · Cited by 2
“However, single LLMs inevitably suffer inherent model biases [3, 18] and weak alignment with nuanced human judgment [1].”
The Epistemic Shift in Science Communication: Strategic Foresight for the Future of Peer Review
Biele, C., & Kowalski, J. (2026) · Science Communication · doi:10.1177/10755470261426817 · Cited by 2
Levchenko, M. (2025) · Anthology of Computers and the Humanities · doi:10.63744/skozvuhqbte7 · Cited by 2
Dycke, N., & Gurevych, I. (2025) · arXiv · doi:10.48550/arxiv.2508.21422 · Cited by 1
Skaar, Ø. O. (2024) · Nordic Journal of Digital Literacy · doi:10.18261/njdl.19.2.4 · Cited by 3
Evaluation of LLMs Biases towards Elite Universities: A Persona-Based Exploration
Ranjan, R., & Gupta, S. (2024) · The Review of Contemporary Scientific and Academic Studies · doi:10.55454/rcsas.4.07.2024.006 · Cited by 18
“In the paper, Gallegos et al. 2024 [4] share the work emphasizing the need for a structured understanding of social bias and fairness in natural language processing.”
BackgroundMarxistLLM: Fine-tuning a language model with a Marxist worldview
Nelimarkka, M. (2026) · Big Data & Society · doi:10.1177/20539517261447831 · Cited by 1
Pentland, S. J., Wang, X., & Twyman, N. W. (2024) · Information & Management · doi:10.1016/j.im.2024.104077 · Cited by 3
Saliency-Based Token Swap - A Language-Agnostic Data Augmentation Method for Text Classification
Ilangeshwaran, H., Abeywardhana, L., & Rathnayake, S. (2024) · International Conference on Information Technology Research (ICITR) · doi:10.1109/icitr64794.2024.10857798 · Cited by 2
Detecting implicit biases of large language models with Bayesian hypothesis testing
Si, S., Jiang, X., Su, Q., & Carin, L. (2025) · Scientific Reports · doi:10.1038/s41598-025-95825-x · Cited by 17
What should philosophers do with “deceptive” intuition pumps? Restrictionism vs reformism
Matsuda, A., & Takeshita, M. (2025) · Philosophical Psychology · doi:10.1080/09515089.2025.2498745 · Cited by 1
“That is, even in the face of deceptive intuition pumps, by relying on the intuitions of artificial intelligences, especially large language models (LLMs), we may be able to determine which of the conflicting intuitions should be prioritized.”
InfluentialEmpowering Future Spectrum Management and Regulation With Large Language Models
Rutagemwa, H., Ghasemi, A., & Guinand, P. S. (2025) · IEEE Intelligent Systems · doi:10.1109/mis.2025.3567372 · Cited by 1
Computational challenges arising in algorithmic fairness and health equity with generative AI
Suriyakumar, V. M., Zink, A., Hightower, M., Ghassemi, M., & Beaulieu‐Jones, B. K. (2025) · Nature Computational Science · doi:10.1038/s43588-025-00806-9 · Cited by 4
ARCHITECTING RESPONSIBLE DEVELOPMENT AND DEPLOYMENT OF GENERATIVE AI
Bhuvaneswari, U., & Prasad, A. (2025) · INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN ENGINEERING & TECHNOLOGY · doi:10.34218/ijaret_16_03_005 · Cited by 1
Muñoz-García, V., Consuegra‐Ayala, J. P., & Moreda, P. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113915 · Cited by 2
WATER: A Two-Stage in-Context Learning Debiasing Framework for Multilingual Text Classification
Huang, W., Long, Z., Zhou, D. M., Chen, Z., Zhou, Y., Lin, N., & Yang, A. (2025) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd64889.2025.11033553 · Cited by 1
“As these inherent biases become more apparent, concerns grow regarding the adverse societal impacts of the content generated by LLMs [5].”
Evaluating the Effects of Prompt Perturbation on Bias and Hallucination in Large Language Models
Yousefi, M., Shahi, A., Sharifi, M., Romera, A. J. J., Hoermann, S., & Piumsomboon, T. (2025) · Lecture notes in computer science · doi:10.1007/978-981-96-6588-4_25 · Cited by 1
“Bias in language models is subjective and shaped by cultural and contextual elements [22].”
Sarhan, H., Shahrezaye, M., & Hegelich, S. (2025) · AI and Ethics · doi:10.1007/s43681-025-00773-x · Cited by 3
Bias and Fairness in Conversational User Interfaces
Laban, G., Hough, J., Lee, M., Markelius, A., Foster, M. E., Stuart‐Smith, J., & Ahmad, M. (2025) · ACM Conference on Conversational User Interfaces (CUI) · doi:10.1145/3719160.3728629 · Cited by 4
Large Language Models for Automated Network Protocol Testing: A Survey
Nguyễn, V. Đ., & Amir-Mohammadian, S. (2025) · IEEE World AI IoT Congress (AIIoT) · doi:10.1109/aiiot65859.2025.11105345 · Cited by 1
Identification of Important LLM Test Criteria for State Media Authorities
Schaffer, S., Kleer, N., Lessel, P., Feld, M., Bauer, J. A., & Goedert, I. (2025) · Mensch und Computer (MuC) · doi:10.1145/3743049.3748541 · Cited by 1
Anomaly Detection in GeoLLM Using Variational Autoencoder
Maladhy, D., & Kumari, K. S. (2025) · International Conference on Data Intelligence and Cognitive Informatics (ICDICI) · doi:10.1109/icdici66477.2025.11135151 · Cited by 1
“In [4], the dataset bias issue arising in the LLM model are discussed in three categories like metrics such as token based embeddings, probabilities and the output text, structure base dataset biasing due to social issues and classification of datasets based on processing in the level of training,…”
Oberbichler, S., Mauermann, J., Tran, T. T., & González-Gallardo, C.-E. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05409-8_16 · Cited by 2
LLMs as annotators: the effect of party cues on labelling decisions by large language models
Vera, S. V., & Driggers, H. (2025) · Humanities and Social Sciences Communications · doi:10.1057/s41599-025-05834-4 · Cited by 17
“More broadly, studies on LLMs have consistently found biases based on contextual and cultural factors (Gallegos et al. 2024), leading to the misrepresentation of certain social groups (Yang et al. 2022), gender stereotyping (Dong et al. 2024), and the reinforcement of normativity (Bender et al.…”
Shafqat, Z., Jilani, A. A. A., Butt, N. A., Rehman, S. U., Khalid, M. U., & Gruhn, V. (2025) · IEEE Access · doi:10.1109/access.2025.3618868 · Cited by 2
“…prior work in conversational bias measurement and LLM fairness evaluation, we adopt Preference Rate (PR) to quantify group-level favorability [2], [23], and Bias Detection Rate (BDR) as a generation-based metric capturing how often bias is elicited across demographic variants [24], [25], [26].”
Against AI Jurisprudence: Large Language Models and the False Promises of Empirical Judging
Pruss, D., & Allen, J. (2025) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1609/aies.v8i3.36695 · Cited by 7
Understanding Generative AI Risks for Youth: An Empirical Taxonomy for Safer Digital Futures
Yu, Y., Liu, Y., Zhang, J., Qian, Z., Huang, Y. M., & Yang, W. (2025) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1526 · Cited by 3
“Recent work on AI ethics and safety has examined broader risks including algorithmic bias, toxicity, misinformation, and privacy violations associated with large language models (Ferrer et al., 2021; Gallegos et al., 2024).”
Can Role-Based LLM Dialogue Improve Understanding of Accessibility Issues?
Miyata, A. (2025) · International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/3663547.3759686 · Cited by 1
Using BERT-like Language Models for Automated Discourse Coding: A Primer and Tutorial
López‐Pernas, S., Misiejuk, K., & Saqr, M. (2025) · Springer eBooks · doi:10.1007/978-3-031-95365-1_10 · Cited by 3
Rewriting Stories with LLMs: Gender Bias in Generated Portuguese-language Narratives
Silva, M. O., Brandão, M. A., & Moro, M. M. (2025) · Journal of the Brazilian Computer Society · doi:10.5753/jbcs.2025.5799 · Cited by 1
“While gender bias in LLMs has been widely studied in English and other high-resource languages [Navigli et al ., 2023; Gallegos et al ., 2024; Ding et al ., 2025], its presence in Portuguese-language models remains underexplored [Lima and Araujo, 2023].”
Mining Impersonification Bias in LLMs via Survey Filling
Bombieri, M., & Rospocher, M. (2025) · Information · doi:10.3390/info16110931 · Cited by 3
“Growing recognition of these challenges has spurred extensive research aimed at assessing and mitigating bias, leading to the creation of evaluation metrics, benchmark datasets, and intervention strategies [1].”
InfluentialNonlinear transformation of probabilities by large language models
Hintze, A., Bisht, C., Schossau, J., & Hertwig, R. (2025) · Computers in Human Behavior: Artificial Humans · doi:10.1016/j.chbah.2025.100227 · Cited by 1
Chatbot for the Preservation of Palestine’s History
Miftah, K., Matrane, Y., Ellaky, Z., & Benabbou, F. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-032-02312-4_32 · Cited by 1
Homophily-induced emergence of biased structures in LLM-based multi-agent AI systems
Mehdizadeh, A., & Hilbert, M. (2025) · Social Network Analysis and Mining · doi:10.1007/s13278-025-01535-7 · Cited by 17
“Recent research has also begun to explore how large language models (LLMs) internalize and reproduce political ideologies and biases present in their training data [111, 112, 113, 114].”
Xu, S., Luo, F., Li, S., Fan, M., & Sun, Z. (2025) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.2500421 · Cited by 2
Exploring cultural commonsense in multilingual large language models: A survey
Binegde, G. N., & Zhang, H. (2025) · Information Systems · doi:10.1016/j.is.2025.102649 · Cited by 3
Goethals, S., Rhue, L. A., & Sundararajan, A. (2025) · AI and Ethics · doi:10.1007/s43681-025-00876-5 · Cited by 4
Emotion in Context: Human and AI Perspectives on the Affective Landscape of Severance
Ortiz, A. M., & García-Gámez, M. (2025) · Corpus Pragmatics · doi:10.1007/s41701-025-00220-6 · Cited by 8
“Thirdly, social biases persist systematically across LLMs, reflecting and often amplifying biases present in training data (Gallegos et al., 2024 ).”
Özçelik, N. P. (2025) · British Educational Research Journal · doi:10.1002/berj.70086 · Cited by 4
RetractedReligion and Racial Bias in Artificial Intelligence Large Language Models
Tom, J. C., Ferguson, T. W., & Martinez, B. C. (2025) · Socius Sociological Research for a Dynamic World · doi:10.1177/23780231251377210 · Cited by 5
“Much in the same way that biases in algorithms reflect assumptions built in from their creators, LLMs have been shown to replicate preexisting social biases on the basis of their training data (Benjamin 2019; Gallegos et al. 2024; Kotek, Dockum, and Sun 2023).”
Template-Driven Specification of Requirements for LLM-Based Chatbots
da Silva, C. V. M., Maita, A. R. C., Filho, G. A. D. A., da Silva, C. H. C., Dimant, G., Almeida, R. A. R., Costa, G. M. L., Araujo, A. M. C., Santana, B. C. D. S., Blay, E. A., et al. (2025) · doi:10.5753/wbots.2025.14151 · Cited by 1
“Recent studies by [Solomon et al. 2024, Cabrera et al. 2024, Gallegos et al. 2024] highlight the need for verification, traceability, and alignment to ensure predictable, compliant, and user-centered behavior.”
When Voice Matters: Evidence of Gender Disparity in Positional Bias of SpeechLLMs
Satish, S., Henter, G. E., & Székely, É. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-07956-5_2 · Cited by 8
“The problem of bias in language modelling and machine learning, particularly with the use of large-scale datasets, has been known and studied for a number of years, with several efforts made to measure and mitigate bias in large language models (LLMs) [2, 4, 20, 28, 10].”
Researching Health Information Behaviors: Landscape, AI 's Role and Its Impact
Yu, X., Zhang, Y., Choi, W., & Chen, A. T. (2025) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1390 · Cited by 1
“Notably, studies have shown that Large Language Models (LLMs) exhibit social biases (e.g., racial, gender, or socioeconomic biases) (Kaneko et al., 2022; Gallegos et al., 2024).”
Liu, S. (2025) · Journal of Computer Science and Artificial Intelligence · doi:10.54097/kptt9j67 · Cited by 2
“The explosive growth in the number of large language models (LLMs) has been a defining shift in natural language processing, enabling unprecedented capabilities for text generation, comprehension, and human-computer interface [1].”
Casting criminals: a framework for evaluating demographic bias in AI-generated narratives
Sheikhi, G. (2025) · Journal of Intelligent Information Systems · doi:10.1007/s10844-025-01016-5 · Cited by 1
Toward human-centered AI-assisted terminology work
Martín, A. S. (2026) · Terminology International Journal of Theoretical and Applied Issues in Specialized Communication · doi:10.1075/term.00095.mar · Cited by 2
“…representations, and discriminatory attitudes arising from historical and structural power asymmetries and directed toward specific social groups based on factors such as gender, race, ethnicity, age, sexual orientation, and disability (Gallegos et al. 2024, 1098; Navigli et al. 2023, 7).”
Large AI Models Have a Prioritization Problem: Policy Implications and Solutions
Jackson, J. C., Liu, Y., Wang, Z., & Brady, W. J. (2025) · Policy Insights from the Behavioral and Brain Sciences · doi:10.1177/23727322251408311 · Cited by 2
“…evaluation criteria beyond traditional accuracy measures to include metrics that capture model fairness and robustness (Agarwal et al., 2018; Gallegos et al., 2024); (3) Interpretable and transparent modeling: develop AI models that provide transparent and comprehensive explanations for…”
Alsayed, A., Elhag, S., & Badri, S. K. (2026) · Information · doi:10.3390/info17010076 · Cited by 1
Chand, S., Baca, F., & Ferrara, E. (2026) · AI · doi:10.3390/ai7010024 · Cited by 9
“However, their increasing adoption is accompanied by a fundamental challenge: LLMs trained on large corpora of human-generated content inherit and frequently exacerbate deeply ingrained societal prejudices regarding race, gender, religion and other sensitive categories [43, 13].”
A Close Reading Approach to Gender Narrative Biases in AI-Generated Stories
Raffini, D., Macori, A., Angelini, M., & Catarci, T. (2025) · IEEE International Conference on Cyber Humanities (IEEE-CH) · doi:10.1109/ieee-ch65308.2025.11279292 · Cited by 3
“Despite ongoing efforts and improvements, LLMs still often do not adequately represent diversity and continue to propagate various forms of societal bias in their output [1] [2] [3].”
Ahmed, H., & Sadka, A. H. (2026) · IEEE Access · doi:10.1109/access.2026.3656396 · Cited by 2
“Advances such as retrieval-augmented generation (RAG) architectures have improved the contextual grounding and trustworthiness of clinical responses [20], yet systematic evaluations, such as those using the AMQA benchmark, high-light persistent disparities in diagnostic accuracy across race, sex,…”
Benchmarking ESG Risk Classification in 10-K Filings
Vo, A., Kim, R., Plachkinova, M., & Lestyk, J. T. (2026) · Journal of Computer Information Systems · doi:10.1080/08874417.2026.2619941 · Cited by 3
Degiovanni, R., & Cabot, J. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00313 · Cited by 1
“Although promising, LLMs are prone to hallucinations [10] and biases [7], which constitutes a big risk for companies aiming to integrate LLMs into their educational products.”
Investigating transformer models for textual bias detection in model, data, and dataspace cards
Donald, A., Galanopoulos, A., Ojha, A. K., Curry, E., Muñoz, E., Ullah, I., McCrae, J. P., Kalra, M., Saxena, S., & Iqbal, T. (2026) · AI and Ethics · doi:10.1007/s43681-025-00975-3 · Cited by 1
“Furthermore, previous studies [7–9] have examined the internal activations of transformer models to visualise and reduce hidden biases, leading to fairness-aware pruning techniques.”
Large Language Model Jailbreak Based on Stylistic Rewriting
Hou, W., Hu, N., Hu, Y., & Luo, W. (2025) · IEEE International Conference on Data Science in Cyberspace · doi:10.1109/dsc67331.2025.00068 · Cited by 2
“…LLMs through specific methods and forcing them to generate content that is prohibited initially or restricted, including violent crime guidance [12], hate speech, personal privacy information [13], [14], prejudice and discriminatory content [15], [16], and content that violates ethics or law [17].”
Malik, S. S., Majid, S., Altaf, S., & Shah, S. A. (2026) · Advanced technologies and societal change · doi:10.1007/978-981-95-0800-6_12 · Cited by 1
Enhancing RLHF in LLMs: Comparing BERT, XGBoost, and Deep Models for Bias Detection
Torfa, W., Elgendy, O., & Nassif, A. B. (2025) · International Conference on Developments in eSystems Engineering (DeSE) · doi:10.1109/dese68208.2025.11367955 · Cited by 2
“These models increasingly deployed in high-impact systems—from chatbots and virtual assistants to hiring tools and healthcare decision support—carry with them the patterns and prejudices embedded in their training data and optimization choices [4][5].”
ReFaRAG: Re-ranking for Bias Mitigation in Retrieval-Augmented Generation
Zhao, Y., Efthymiou, V., Nummenmaa, J., & Stefanidis, K. (2025) · Communications in computer and information science · doi:10.1007/978-3-032-05727-3_42 · Cited by 7
Rizqullah, M., & Albassam, E. (2026) · IEEE Access · doi:10.1109/access.2026.3662817 · Cited by 4
“There have also been studies that proved LLMs to be biased or produce harmful content [12], [13].”
AI-Driven Digital Credentials: Building a Smarter Credential Outcome Recognition Ecosystem
Lin-Cook, W. W.-C., & Chun, D. (2026) · Educational communications and technology: issues and innovations · doi:10.1007/978-3-032-06823-1_9 · Cited by 1
Tanvir, M., Akter, A., Ahammad, M. S., Tahmid, M., Sraboni, S. A., & Islam, T. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-15346-3_2 · Cited by 1
Interviewing AI-Generated Personas: Talking To Your Data to Generate Qualitative Text from Users
Azem, J. Y., Ahmed, F., Salminen, J. O., & Jansen, B. J. (2025) · International Conference on Foundation and Large Language Models (FLLM) · doi:10.1109/fllm67465.2025.11390880 · Cited by 3
“LLMs are known to have biases reflecting their training corpora [32], and if not prompted, they might default to stereotypical or majority viewpoints [11].”
Automated test plan generation using large language models
Haldar, S., & Capretz, L. F. (2026) · Software Quality Journal · doi:10.1007/s11219-026-09744-9 · Cited by 3
“These concerns are also consistent with Gallegos et al. (2024 ), who surveyed bias and fairness in LLMs, and Zhao et al. ( 2026), who reviewed retrieval augmented generation as a strategy to improve reliability.”
Reframing the Performance and Ethics of Empathic AI: Wisdom of the Crowd and Placebos
Thornton, M. A. (2026) · Perspectives on Psychological Science · doi:10.1177/17456916261419718 · Cited by 1
“Interests and biases shared by all or many people in the training data—such as prejudices against marginalized social groups—are likely to persist or even be amplified by AI unless targeted remediation is undertaken (Caliskan et al., 2017; Gallegos et al., 2023; Nazer et al., 2023).”
Fair In-Context Learning via Latent Concept Variables
Bhaila, K., Van, M.-H., Edemacu, K., Zhao, C., Chen, F., & Wu, X. (2025) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata66926.2025.11402286 · Cited by 7
“Multiple works have highlighted discriminatory behavior in LLM outputs originating from societal biases contained in pre-training data [6], [12] or under-representation of minority population [13].”
Khan, S. H., Williams, M. S., & Lybarger, K. (2026) · PLOS Digital Health · doi:10.1371/journal.pdig.0001275 · Cited by 1
“Pre-trained language models may reflect biases present in large-scale training data, including social, cultural, and demographic prejudices [56,57].”
Xiong, X., Xiao, M., Liu, H., Jiang, C., Li, M., Song, Y., Zhou, J., & Xu, B. (2026) · BMC Medical Education · doi:10.1186/s12909-026-08951-z · Cited by 2
“Another study provided a comprehensive survey of bias evaluation and mitigation techniques for LLMs, offering insights into how to identify and address these biases through various strategies [19].”
Rad, A.-C., & Lemnaru, C. (2025) · IEEE International Conference on Intelligent Computer Communication and Processing (ICCP) · doi:10.1109/iccp68926.2025.11427185 · Cited by 1
“Gallegos et al. [11] have provided taxonomies of bias identification and mitigation techniques for LLMs, yet important gaps remain, especially around multiculturalism and multilingualism.”
Ahmadi, A., Baudry, J. Y., Tenhundfeld, N. L., Goff, K., & Adamek, D. (2026) · PLoS ONE · doi:10.1371/journal.pone.0342700 · Cited by 1
“From a regulatory standpoint, ensuring model transparency, explainability, and human-in-the-loop review processes will be necessary for ethical compliance and user trust [73].”
The 9th International Workshop on Narrative Extraction from Text: Text2Story 2026
Campos, R., Jorge, A. O. S., Jatowt, A., Bhatia, S., & Litvak, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-21324-2_10 · Cited by 1
Toxicbias-reasoning: a multicultural dataset for social bias detection with human-aligned reasoning
Kumar, A., Gurve, M. K., Ahlawat, S., Prasad, Y., & Singh, V. (2026) · Language Resources and Evaluation · doi:10.1007/s10579-026-09916-w · Cited by 1
“This amplification of harmful patterns can result in outputs that legitimise hate speech, normalise derogatory language, and reinforce stereotypes, thereby raising critical concerns about fairness, accountability, and ethical deployment of AI systems (Gallegos et al., 2024).”
A review of fairness challenges in natural language processing
Zadid, T. H., Jahin, S. M., Dhruba, A. R., Tanim, S. A., Hamja, A. M., Hasan, M. R., Miah, M. S. U., Islam, M. I., & Mansur, A. A. (2026) · Array · doi:10.1016/j.array.2026.100773 · Cited by 4
Bombieri, M., Ponzetto, S. P., & Rospocher, M. (2026) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3806202 · Cited by 2
Jeon, S., Lee, M. H. J., Montgomery, J., & Lai, C. K. (2026) · Political Analysis · doi:10.1017/pan.2026.10038 · Cited by 1
Foundation Models for Autonomous Driving Systems: An Initial Roadmap
Wu, X., Cheng, M., Ren, X., Hu, Q., Chen, J., Huang, Y., Cordy, M., Zhang, Y., Xie, X., Ma, L., et al. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3807455 · Cited by 18
Ask ChatGPT: caveats and mitigations for individual users of AI chatbots
Wang, C., & Kantarcıoğlu, M. (2026) · AI and Ethics · doi:10.1007/s43681-026-01089-0 · Cited by 3
Darabipourshiraz, H., Smyslova, D., Xu, D., Jiang, S., & Long, D. (2026) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791555 · Cited by 3
Characterizing User-Reported Risks across LLM Chatbots
Li, L., Ma, R., Xue, Z., & Xiong, J. (2026) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791505 · Cited by 3
Requirement Classification using Transfer Learning Models
Alsanoosy, T. (2026) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-026-11242-w · Cited by 1
“The deployment of transformer-based models and LLMs raises ethical implications regarding systemic bias, data privacy, fairness and data provenance [98, 99].”
Edin, Y., Rosenplänter, L., & Kersting, M. (2026) · Psychological Test Adaptation and Development · doi:10.1027/2698-1866/a000127 · Cited by 1
Hegberg, A., Setor, T. K., & Joseph, D. (2026) · ACM SIGCPR Conference on Computer Personnel Research (SIGCPR) · doi:10.1145/3768310.3807790 · Cited by 1
“This means scanning demographic narratives for recurring scripts and revising personas or regenerating them with counter-stereotypical constraints so that requirements are not anchored to a narrow social storyline (Abid et al., 2021; Blodgett et al., 2020; Gallegos et al., 2024).”
InfluentialLee, H., Seo, S., Park, S., Tam, L., Kim, S., & Jang, Y. (2026) · Public Relations Review · doi:10.1016/j.pubrev.2026.102709 · Cited by 1
Self-citationSelective Persuasion: Structural Stance Bias of Generative AI in Controversial Issues
黎, 樟., & 塔, 娜. (2026) · Global Journal of Media Studies · doi:10.26599/gjms.2026.9330009 · Cited by 1
Debiasing and Explainability Effectiveness of LLM Tuning: A Quantitative Approach
Mauri, L., Preda, F., & Damiani, E. (2026) · IEEE Conference on Artificial Intelligence (CAI) · doi:10.1109/cai68641.2026.11536280 · Cited by 2
“Most LLM bias metrics operate as aggregate scores over large template collections and therefore do not decompose cleanly at the token level [33].”
Guetterman, T. C., James, T. G., & DeJonckheere, M. J. (2026) · Methods in Psychology · doi:10.1016/j.metip.2026.100266 · Cited by 1
BIPOLAR: Polarization-based granular framework for LLM bias evaluation
Filip, T., Pavlíček, M., & Sosík, P. (2026) · Theoretical Computer Science · doi:10.1016/j.tcs.2026.116117 · Cited by 1
The case for systematic reviews in religion: an under-used way of approaching previous research
Larsson, G., & Eidrup, M. (2026) · Religion · doi:10.1080/0048721x.2026.2690653 · Cited by 2
Sato, K., Niimi, Y., Ihara, R., Suzuki, K., Iwata, A., & Iwatsubo, T. (2026) · Journal of Alzheimer s Disease · doi:10.1177/13872877261464182 · Cited by 1
“20 This can lead to the ampli fi cation of existing societal biases, 21 the under-representation of minority viewpoints, 22 and responses that are plausible on the surface but lack genuine human grounding (hallucination).”
Suherman, A., Maulana, H. F., & Hasrullah, H. (2026) · Journal of Information Communication and Ethics in Society · doi:10.1108/jices-07-2025-0162 · Cited by 1
Bougzime, O., Cruz, C., Zhou, K., Qi, H. J., & Demoly, F. (2026) · Computers in Industry · doi:10.1016/j.compind.2026.104522 · Cited by 4
Welch, C. J., Dunbar, S. B., & Vo, Y. (2026) · Assessing Writing · doi:10.1016/j.asw.2026.101107 · Cited by 3
Fang, Q., & Liu, W. (2026) · Computers and Education Artificial Intelligence · doi:10.1016/j.caeai.2026.100664 · Cited by 2
Peykani, P., Charles, V., Emrouznejad, A., Ghanidel, S., & Sabour, D. (2026) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-026-10720-6 · Cited by 1
Yeung, S., & Gray, J. (2026) · TESOL Quarterly · doi:10.1002/tesq.70208 · Cited by 1
Durán, A. G., de Oliver-Lopez, C. P., & Bravo, C. B. (2026) · Journal of Cultural Heritage · doi:10.1016/j.culher.2026.08.010 · Cited by 1
System level governance of generative AI risks in cross border e-commerce supply chains
Bostan, I. (2026) · Discover Computing · doi:10.1007/s10791-026-10509-y · Cited by 2
“The literature documents how biased training data and model architectures can distort demand forecasting, dynamic pricing, supplier evaluation, and credit scoring in digital marketplaces [5, 10, 11, 56].”
Uncertainty-aware large language models: a scoping review of conformal prediction methods
Ashby, A. E., Nguyen, K. A., & Luo, Z. (2026) · Philosophical Transactions of the Royal Society A Mathematical Physical and Engineering Sciences · doi:10.1098/rsta.2026.0061 · Cited by 1
Villarreal‐Zegarra, D., Paredes-Gonzales, Y., & García-Serna, J. (2026) · PLOS mental health. · doi:10.1371/journal.pmen.0000702 · Cited by 1
“…or DIF of the items or indicators; differential performance of scoring algorithms; subgroup analysis of missingness and data quality; and audit of generated feedback or reports for stereotyping, stigmatizing language, unsafe recommendations, or unequal uncertainty communication [31,33,34].”
InfluentialAgainst mandatory prompt disclosure in AI-assisted scholarship
Earp, B. D., Schüklenk, U., Savulescu, J., & Mann, S. P. (2026) · AI & Society · doi:10.1007/s00146-026-03326-w · Cited by 1
Medical AI in the EU: Regulatory Considerations and Future Outlook
Ranttila, P., Sahebi, G., Kontio, E., & Salmi, J. (2024) · Artificial intelligence · doi:10.5772/intechopen.1007443 · Cited by 2
Metadata
- Authors
- Gallegos, I. O., Rossi, R. A., Barrow, J., Tanjim, M. M., Kim, S., Dernoncourt, F., Yu, T., Zhang, R., & Ahmed, N. K.
- Affiliations
- Stanford University (US)
- Adobe Research. ryrossi@adobe.com
- Pattern Data. joe.barrow@patterndataworks.com
- Adobe Research. tanjim@adobe.com
- Adobe Research. sukim@adobe.com
- Adobe Research. dernonco@adobe.com
- Adobe Research. tyu@adobe.com
- Adobe Research. ruizhang@adobe.com
- Intel (United States) (US)
- ORCID
- Venue
- Computational Linguistics
- Publisher
- MIT Press
- Published
- 2024
- Volume
- 50
- Issue
- 3
- Pages
- 1097-1179
- ISSN
- 0891-2017 (ISSN-L)
- 1530-9312
- OpenAlex
- W4399528455
- Semantic Scholar
- bcfa73aedf1b2d1ee4f168e21298a37ac55a37f7
- Cited by
- 1,245
- References
- 280
- Citations by year
- 2026: 596 · 2025: 451 · 2024: 188
- FWCI
- 132.76
- Type
- journal article
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- Locations
- Computational Linguistics (publishedVersion, cc-by-nc-nd)
- arXiv (Cornell University) (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- large language models, social bias, fairness, bias assessment, bias mitigation, bias metrics, data preprocessing, post-processing
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-11-02, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-02 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.