CoChat
Sign Up
Cited by 439Decision-Making and Behavioral EconomicsNeural and Behavioral Psychology StudiesNeurotransmitter Receptor Influence on Behavior

The DOI matches its Crossref record; 69 of the 70 references it cites were found in a registry or an index; none of the 29 sources its claims cite is checked for support yet.

Bonsai Trees in Your Head: How the Pavlovian System Sculpts Goal-Directed Choices by Pruning Decision Trees

Open access
Huys, Quentin J. M., Eshel, Neir, O'Nions, Elizabeth, Sheridan, Luke, Dayan, Peter, Roiser, Jonathan P.·2012·PLoS Computational Biology
DOI 10.1371/journal.pcbi.1002410PMC PMC3297555
Open this paper in CoChat — free accountOpen PDF

Abstract

When planning a series of actions, it is usually infeasible to consider all potential future sequences; instead, one must prune the decision tree. Provably optimal pruning is, however, still computationally ruinous and the specific approximations humans employ remain unknown. We designed a new sequential reinforcement-based task and showed that human subjects adopted a simple pruning strategy: during mental evaluation of a sequence of choices, they curtailed any further evaluation of a sequence as soon as they encountered a large loss. This pruning strategy was Pavlovian: it was reflexively evoked by large losses and persisted even when overwhelmingly counterproductive. It was also evident above and beyond loss aversion. We found that the tendency towards Pavlovian pruning was selectively predicted by the degree to which subjects exhibited sub-clinical mood disturbance, in accordance with theories that ascribe Pavlovian behavioural inhibition, via serotonin, a role in mood disorders. We conclude that Pavlovian behavioural inhibition shapes highly flexible, goal-directed choices in a manner that may be important for theories of decision-making in mood disorders.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

Humans planning sequences prune decision trees by curtailing evaluation after large losses, a Pavlovian response that persists even when counterproductive. Computational modeling shows this specific pruning is distinct from general discounting and loss aversion, and correlates with subclinical depressive symptoms.

Key findings

  • Subjects adopted a simple pruning strategy: they stopped evaluating a sequence when they encountered a large loss, even when this was costly. In group −140 pruning was never disadvantageous, but in groups −70 and −100 it was increasingly counterproductive, yet subjects continued to prune.
  • The most parsimonious model was “Pruning & Learned,” which separated general pruning from specific pruning after large punishments. Specific pruning exceeded general pruning in 14/15, 12/16, and 14/15 subjects in the −70, −100, and −140 groups, respectively. This model correctly predicted 91% of choices; the full look-ahead model predicted 77% of choices and had a mean predictive probability of 0.71.
  • Pruning was evident above and beyond loss aversion. Subjects were approximately 1.4 times more sensitive to punishments than to equal-sized rewards, but adding loss aversion to the best model did not improve parsimony.
  • BDI score positively correlated with the specific pruning parameter, but not with general pruning. A negative correlation between BDI and reward sensitivity did not survive correction for multiple comparisons. NEO agreeableness positively correlated with the Learned Pavlovian influence and survived full correction for 60 comparisons; no significant correlation with NEO neuroticism was found.

Methods

  • Forty-six volunteers (23 female, mean age 23.84 years) were recruited and screened; groups received different large losses: −70 (n=15), −100 (n=16), and −140 (n=15). Subjects learned a deterministic six-state transition matrix, then completed two blocks of 24 choice episodes with sequence lengths of 2–8; the first block was discarded and the second analyzed. On half the trials, subjects planned their last 2–4 moves together without intermittent feedback.
  • Computational models included full look-ahead, discounting, pruning (general and specific), learned Pavlovian attraction/repulsion, loss aversion, and a habitual SARSA model. Models were fit with Bayesian model fitting and compared using integrated Bayesian Information Criterion at the group level.
  • Psychometric correlates were assessed with a weighted hierarchical multivariate regression, orthogonalizing age, sex, IQ, Digit Span, NEO measures, STAI, and BDI.

Limitations

  • The paper states that the sample was healthy volunteers with only very mild depressive symptoms, so the BDI correlation cannot be taken as direct evidence that mildly raised BDI scores measure vulnerability to depression; predictions about clinical depression and serotonin depletion remain to be tested. It also states that no significant correlation with NEO neuroticism was found.
  • The paper notes that in group −140 there was always an optimal sequence avoiding the large loss, making that group’s behavioral statistic difficult to interpret; at depths of 6 or more, the number of events was small, limiting power; and some missing age/STAI data excluded subjects from psychometric analyses.
  • Evident from the text: the task uses a fixed deterministic transition matrix and specific reinforcement structure, and the conclusions depend on the fitted model space; model comparison did not support separate group differences in pruning despite apparent behavioral differences.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pcbi.1002410

Cites

69 of 69 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Decision tree. A: A typical decision tree. A sequence of choices between ‘U’ (left, green) and ‘I’ (right, orange) is made to maximize the total amount earned over the entire sequence of choices. Two sequences yield the maximal total outcome of −20 (three times U; or I then twice U). Finding the optimal choice in a goal-directed manner requires evaluating all 8 sequences of three moves each. B: Pruning a decision tree at the large negative outcome. In this simple case, pruning would still favour one of the two optimal sequences (yielding −20), yet cut the computational cost by nearly half.
Figure 1 Decision tree. A: A typical decision tree. A sequence of choices between ‘U’ (left, green) and ‘I’ (right, orange) is made to maximize the total amount earned over the entire sequence of choices. Two sequences yield the maximal total outcome of −20 (three times U; or I then twice U). Finding the optimal choice in a goal-directed manner requires evaluating all 8 sequences of three moves each. B: Pruning a decision tree at the large negative outcome. In this simple case, pruning would still favour one of the two optimal sequences (yielding −20), yet cut the computational cost by nearly half.CC BY · PubMed Central
Figure 2 Task description. A: Task as seen by subjects. Subjects used two buttons on the keyboard (‘U’ and ‘I’) to navigate between six environmental states, depicted as boxes on a computer screen. From each state, subjects could move to exactly two other states. Each of these was associated with a particular reinforcement. The current state was highlighted in white, and the required sequence length displayed centrally. Reinforcements available from each state were displayed symbolically below the state, e.g. for the large reward. B: Deterministic task transition matrix. Each button resulted in one of two deterministic transitions from each state. For example, if the participant began in state 6, pressing ‘U’ would lead to state 3, whereas pressing ‘I’ would lead to state 1. The transitions in red yielded large punishments. These (and only these) differed between three groups of subjects (−140, −100 or −70). Note that the decision trees in Figure 1A,B correspond to a depth 3 search starting from state 3. C–E: Effect of pruning on values of optimal choices. Each square in each panel analyses choices from one state when a certain number of choices remains to be taken. The color shows the difference in earnings between two choice sequences: the best choice sequence with pruning and the best choice sequence without pruning. In terms of net earnings, pruning is never advantageous (pruned values are never better than the optimal lookahead values); but pruning does not always result in losses (white areas). It is most disadvantageous in the −70 group, and it is never disadvantageous in the −140 group because there is always an equally good alternative choice sequence which avoids transitions through large losses.
Figure 2 Task description. A: Task as seen by subjects. Subjects used two buttons on the keyboard (‘U’ and ‘I’) to navigate between six environmental states, depicted as boxes on a computer screen. From each state, subjects could move to exactly two other states. Each of these was associated with a particular reinforcement. The current state was highlighted in white, and the required sequence length displayed centrally. Reinforcements available from each state were displayed symbolically below the state, e.g. for the large reward. B: Deterministic task transition matrix. Each button resulted in one of two deterministic transitions from each state. For example, if the participant began in state 6, pressing ‘U’ would lead to state 3, whereas pressing ‘I’ would lead to state 1. The transitions in red yielded large punishments. These (and only these) differed between three groups of subjects (−140, −100 or −70). Note that the decision trees in Figure 1A,B correspond to a depth 3 search starting from state 3. C–E: Effect of pruning on values of optimal choices. Each square in each panel analyses choices from one state when a certain number of choices remains to be taken. The color shows the difference in earnings between two choice sequences: the best choice sequence with pruning and the best choice sequence without pruning. In terms of net earnings, pruning is never advantageous (pruned values are never better than the optimal lookahead values); but pruning does not always result in losses (white areas). It is most disadvantageous in the −70 group, and it is never disadvantageous in the −140 group because there is always an equally good alternative choice sequence which avoids transitions through large losses.CC BY · PubMed Central
Figure 3 Choice sequences. Example decision trees of varying depth starting from states 1 or 3. The widths of the solid lines are proportional to the frequencies with which particular paths were chosen (aggregated across all subjects). Yellow backgrounds denote optimal paths (note that there can be multiple optimal paths). Colours red, black, green and blue denote transitions with reinforcements of and respectively. Dashed lines denote parts of the decision tree that were never visited. Visited states are shown in small gray numbers where space allows. A: Subjects avoid transitions through large losses. In the condition, this is not associated with an overall loss. B: In the condition, where large rewards lurk behind the losses, subjects can overcome their reluctance to transition through large losses and can follow the optimal path through an early large loss. C: However, they do this only if the tree is small and thus does not require pruning. Subjects fail to follow the optimal path through the same subtree as in B (indicated by a black box) if it occurs deeper in the tree, i.e. in a situation where computational demands are high. D,E,F Fraction of times subjects in each group chose the optimal sequence, deduced by looking all the way to the end of the tree. Green shows subjects' choices when the optimal sequence did not contain a large loss; blue shows subjects' choices when the optimal sequence did contain a large loss. Coloured areas show 95% confidence intervals, and dashed lines predictions from the model ‘Pruning & Learned’ (see below).
Figure 3 Choice sequences. Example decision trees of varying depth starting from states 1 or 3. The widths of the solid lines are proportional to the frequencies with which particular paths were chosen (aggregated across all subjects). Yellow backgrounds denote optimal paths (note that there can be multiple optimal paths). Colours red, black, green and blue denote transitions with reinforcements of and respectively. Dashed lines denote parts of the decision tree that were never visited. Visited states are shown in small gray numbers where space allows. A: Subjects avoid transitions through large losses. In the condition, this is not associated with an overall loss. B: In the condition, where large rewards lurk behind the losses, subjects can overcome their reluctance to transition through large losses and can follow the optimal path through an early large loss. C: However, they do this only if the tree is small and thus does not require pruning. Subjects fail to follow the optimal path through the same subtree as in B (indicated by a black box) if it occurs deeper in the tree, i.e. in a situation where computational demands are high. D,E,F Fraction of times subjects in each group chose the optimal sequence, deduced by looking all the way to the end of the tree. Green shows subjects' choices when the optimal sequence did not contain a large loss; blue shows subjects' choices when the optimal sequence did contain a large loss. Coloured areas show 95% confidence intervals, and dashed lines predictions from the model ‘Pruning & Learned’ (see below).CC BY · PubMed Central
Figure 4 Model performance and comparison. A: Fraction of choices predicted by the model as a function of the number of choices remaining. For bars ‘3 choices to go’, for instance, it shows the fraction of times the model assigned higher value to the subject's choice in all situations where three choices remained (i.e. bar 3 in these plots encompasses all three panels in Figure 3A–C). These are predictions only in the sense that the model predicts choice based on history up to . The gray line shows this statistic for the full look-ahead model, and the blue bars for the most parsimonious model (‘Pruning and Learned’). B: Mean predictive probabilities, i.e. likelihood afforded to choices on trial given learned values up to trial . C: Model comparison based on integrated Bayesian Information Criterion () scores. The lower the score, the more parsimonious the model fit. For guidance, some likelihood ratios are displayed explicitly, both at the group level (fixed effect) and at the individual level (random effect). Our main guide is the group-level (fixed effect). The red star indicates the most parsimonious model. D,E: Transition probability from state 6 to state 1 (which incurs a −20 loss) when a subsequent move to state 2 is possible (D; at least two moves remain) or not (E; when it is the only remaining move). Note that subjects' disadvantageous approach behavior in E (dark gray bar) is only well accommodated by a model that incorporates the extra Learned Pavlovian parameter. F: Decision tree of depth 4 from starting state 3. See Figure 3 for colour code. Subjects prefer (width of line) the optimal (yellow) path with an early transition through a large loss (red) to an equally optimal path with a late transition through a large loss. G: Phase plane analysis of specific and general pruning. Parameter values for which the left optimal yellow path in panel F is assigned a greater expected value than the right optimal path are below the blue line. Combinations that are also consistent with the notion of pruning are shown in green. The red dot shows parameters inferred for present data (c.f. Figure 6). Throughout, errorbars indicate one standard error of the mean (red) and the 95% confidence intervals (green).
Figure 4 Model performance and comparison. A: Fraction of choices predicted by the model as a function of the number of choices remaining. For bars ‘3 choices to go’, for instance, it shows the fraction of times the model assigned higher value to the subject's choice in all situations where three choices remained (i.e. bar 3 in these plots encompasses all three panels in Figure 3A–C). These are predictions only in the sense that the model predicts choice based on history up to . The gray line shows this statistic for the full look-ahead model, and the blue bars for the most parsimonious model (‘Pruning and Learned’). B: Mean predictive probabilities, i.e. likelihood afforded to choices on trial given learned values up to trial . C: Model comparison based on integrated Bayesian Information Criterion () scores. The lower the score, the more parsimonious the model fit. For guidance, some likelihood ratios are displayed explicitly, both at the group level (fixed effect) and at the individual level (random effect). Our main guide is the group-level (fixed effect). The red star indicates the most parsimonious model. D,E: Transition probability from state 6 to state 1 (which incurs a −20 loss) when a subsequent move to state 2 is possible (D; at least two moves remain) or not (E; when it is the only remaining move). Note that subjects' disadvantageous approach behavior in E (dark gray bar) is only well accommodated by a model that incorporates the extra Learned Pavlovian parameter. F: Decision tree of depth 4 from starting state 3. See Figure 3 for colour code. Subjects prefer (width of line) the optimal (yellow) path with an early transition through a large loss (red) to an equally optimal path with a late transition through a large loss. G: Phase plane analysis of specific and general pruning. Parameter values for which the left optimal yellow path in panel F is assigned a greater expected value than the right optimal path are below the blue line. Combinations that are also consistent with the notion of pruning are shown in green. The red dot shows parameters inferred for present data (c.f. Figure 6). Throughout, errorbars indicate one standard error of the mean (red) and the 95% confidence intervals (green).CC BY · PubMed Central
Figure 5 Pruning exists above and beyond any loss aversion. A: Loss aversion model comparison scores. Red star indicates most parsimonious model. The numbers by the bars show model likelihood ratios of interest at the group level, and below them at the mean individual level. Pruning adds parsimony to the model even after accounting for loss aversion (cf. ‘Discount & Loss’ vs ‘Pruning & Loss’), while loss aversion does not increase parsimony when added to the best previous model (‘Pruning & Learned’ vs ‘Loss & Prune & Learned’). B: Separate inference of all reinforcement sensitivities from best loss aversion model. C: Absolute ratio of inferred sensitivity to maximal punishment (−70, −100 or −140) and inferred sensitivity to maximal reward (always +140). Subjects are 1.4 times more sensitive to punishments than to rewards.
Figure 5 Pruning exists above and beyond any loss aversion. A: Loss aversion model comparison scores. Red star indicates most parsimonious model. The numbers by the bars show model likelihood ratios of interest at the group level, and below them at the mean individual level. Pruning adds parsimony to the model even after accounting for loss aversion (cf. ‘Discount & Loss’ vs ‘Pruning & Loss’), while loss aversion does not increase parsimony when added to the best previous model (‘Pruning & Learned’ vs ‘Loss & Prune & Learned’). B: Separate inference of all reinforcement sensitivities from best loss aversion model. C: Absolute ratio of inferred sensitivity to maximal punishment (−70, −100 or −140) and inferred sensitivity to maximal reward (always +140). Subjects are 1.4 times more sensitive to punishments than to rewards.CC BY · PubMed Central
Figure 6 Pruning parameters. A: Pruning parameter estimates – specific and general pruning parameters are shown separately for each group. Specific pruning exceeded general pruning across subjects, but there was no main effect of group and no interaction. The difference between parameter types was significant in all three groups, with specific exceeding general pruning for 14/15, 12/16 and 14/15 subjects in the −70, −100 and −140 groups respectively. Blue bars show specific pruning parameters () and red bars general pruning parameters (). Black dots show the estimates for each subject. Gray lines show the uncertainty (square root of second moment around the parameter) for each estimate. B: Equivalent parametrization of the most parsimonious model to infer differences between pruning and discount factors directly. For all three groups, the difference is significantly positive. C: Income lost due to pruning. On trials on which the optimal sequence led through large punishments, subjects lost more income the more counterproductive pruning was (loss in group −70loss in group −100loss in group −140). Each bar shows the total income subjects lost because they avoided transitions through large losses. Throughout, the bars show the group means, with one standard error of the mean in red and the 95% confidence interval in green.
Figure 6 Pruning parameters. A: Pruning parameter estimates – specific and general pruning parameters are shown separately for each group. Specific pruning exceeded general pruning across subjects, but there was no main effect of group and no interaction. The difference between parameter types was significant in all three groups, with specific exceeding general pruning for 14/15, 12/16 and 14/15 subjects in the −70, −100 and −140 groups respectively. Blue bars show specific pruning parameters () and red bars general pruning parameters (). Black dots show the estimates for each subject. Gray lines show the uncertainty (square root of second moment around the parameter) for each estimate. B: Equivalent parametrization of the most parsimonious model to infer differences between pruning and discount factors directly. For all three groups, the difference is significantly positive. C: Income lost due to pruning. On trials on which the optimal sequence led through large punishments, subjects lost more income the more counterproductive pruning was (loss in group −70loss in group −100loss in group −140). Each bar shows the total income subjects lost because they avoided transitions through large losses. Throughout, the bars show the group means, with one standard error of the mean in red and the 95% confidence interval in green.CC BY · PubMed Central
Figure 7 Psychometric correlates. A: Subclinical depression scores (Beck Depression Inventory, BDI, range 0–15) correlated positively with specific pruning (), and negatively with sensitivity to the reinforcers (). Each bar shows a weighted linear regression coefficient. Red error bars show one standard error of the mean estimate, and green errorbars the Bonferroni corrected 95% confidence interval. , red dot . B,C: Weighted scatter plots of psychometric scores against parameters after orthogonalization.
Figure 7 Psychometric correlates. A: Subclinical depression scores (Beck Depression Inventory, BDI, range 0–15) correlated positively with specific pruning (), and negatively with sensitivity to the reinforcers (). Each bar shows a weighted linear regression coefficient. Red error bars show one standard error of the mean estimate, and green errorbars the Bonferroni corrected 95% confidence interval. , red dot . B,C: Weighted scatter plots of psychometric scores against parameters after orthogonalization.CC BY · PubMed Central

Verification

Claims the paper makes

35 Own result · 20 Not checked yet · 9 Own interpretation · 6 Own method · 4 No basis given

Verify all 29 references — free account

Show all 41 claims

  • It may state two separate claims as one. 24
  • Probably not a substantial claim. 18
33 of 33

  • The paper relies on this without support.

    The large losses in the task are aversive/punishing events capable of evoking Pavlovian behavioural inhibition.

    “We term this heuristic pruning a “Pavlovian” response because it is invoked, as an immediate consequence of encountering the large loss, when searching the tree in one's mind.”

  • The paper relies on this without support.

    The BDI validly measures sub-clinical depressive symptoms/mood disturbance.

    “We found that BDI was positively correlated with the specific pruning parameter ().”

  • The authors state this as an assumption.

    Actions are independent given the stimuli, and thus factorize over trials.

    “We assume that actions are independent (given the stimuli, which we omit for notational clarity), and thus factorize over trials.”

  • The authors state this as an assumption.

    The value of each choice depends on how many choices are left after it, but not on the choices preceding it.

    “we emphasize that the value of each choice depends on how many choices are left after , but not on the choices preceding it.”

  • The paper relies on this without support.

    The one-touch Tower of London (OTT) task engages goal-directed processing.

    “Contrary to this, there is a recent report that mood manipulation, but not acute tryptophan depletion, impairs processing on the one-touch Tower of London (OTT) task [40], which should certainly engage goal-directed processing.”

  • The paper relies on this without support.

    The pruning parameter mediates the inhibition of iterative thought processes.

    “if serotonin indeed mediates behavioural inhibition in the face of punishments [10], [12]–[14] then it is a strong prediction that the pruning parameter , which mediates the inhibition of iterative thought processes, should be related to, and modulated by, serotonergic activity.”
6 of 6

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?

Citations

Cites

Show all 70 references

Cited by

Sorted by citations

See all 439

Related

Metadata

Authors
Huys, Q. J. M., Eshel, N., O'Nions, E., Sheridan, L., Dayan, P., & Roiser, J. P.
Affiliations
  • Wellcome Centre for Human Neuroimaging (GB)
  • University College London (GB)
  • Gatsby Computational Neuroscience Unit (GB)
Corresponding author
Huys, Quentin J. M.
ORCID
Venue
PLoS Computational Biology
Publisher
Public Library of Science (PLoS)
Published
2012-03-08
Volume
8
Issue
3
Pages
e1002410
ISSN
  • 1553-734X (ISSN-L)
  • 1553-7358
DOI
10.1371/journal.pcbi.1002410
PMC
PMC3297555
OpenAlex
W2059569317
Semantic Scholar
09c66a4231400ba276c92700fe993d6c185bc4ac
Cited by
439
References
70
Citations by year
2026: 16 · 2025: 35 · 2024: 32 · 2023: 42 · 2022: 38 · 2021: 30 · 2020: 33 · 2019: 32 · 2018: 44 · 2017: 34 · 2016: 27 · 2015: 27 · 2014: 29 · 2013: 11 · 2012: 5
FWCI
17.75
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS Computational Biology (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Staff Publications Online (The Tavistock and Portman) (acceptedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • UCL Discovery (University College London) (submittedVersion)
  • MPG.PuRe (Max Planck Society) (submittedVersion)
Funders
Gatsby Charitable Foundation
Topics
  • Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neurotransmitter Receptor Influence on Behavior
Keywords
decision tree pruning, goal-directed behavior, sequential decision making, loss aversion, serotonin, mood disorders, reinforcement learning, human decision-making
MeSH terms
Algorithms, Computer Simulation, Decision Making, Goals, Humans, Models, Neurological, Reinforcement, Psychology, Task Performance and Analysis

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.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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