Not Yet AlphaFold for the Mind: Evaluating Centaur as a Synthetic Participant

Alessandra Brondetta; Matthew Nassar; Sabrina Namazova; Sebastian Musslick; Younes Strittmatter

REVIEW 3 major objections 4 minor 1 cited by

Not Yet AlphaFold for the Mind: Evaluating Centaur as a Synthetic Participant

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Centaur, a language model fine-tuned on 160 human experiments, predicts participants' next choices accurately yet systematically fails to generate human-like behavior when run open-loop, so it does not yet qualify as a reliable participant

desk verdict Useful first open-loop evaluation of Centaur's generative behavior, with a clear predictive-generative dissociation; the reversal-learning benchmark is synthetic RW data, not human, so the 'divergence from human data' framing overreaches on that task. read the letter →

arxiv 2508.07887 v1 pith:HF3JPVER submitted 2025-08-11 cs.LG cs.AI

Sabrina Namazova , Alessandra Brondetta , Younes Strittmatter , Matthew Nassar , Sebastian Musslick This is my paper · ORCID

classification cs.LGcs.AI

keywords participantsimulatorlargelanguagemodelcognitivemodelingpredictivevsgenerativeevaluationreversallearninghorizoneffectWisconsinCardSortingTestfoundationofbehavior

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Centaur is a large language model fine-tuned on human behavioral data from 160 experiments, proposed as a model of cognition and as a synthetic participant for prototyping studies. This paper argues that the standard test used to validate such models—trial-by-trial prediction of the next human choice—cannot certify that the model can generate human-like behavior from scratch. The authors evaluate Centaur in two modes: predictive, where the model sees the human's actual history, and generative, where it sees only its own history. Across a reversal learning task, a horizon-dependent bandit task, and the Wisconsin Card Sorting Test, Centaur's generative behavior systematically misses the qualitative hallmarks these tasks were designed to measure, even when its predictive performance is strong. The paper concludes that predictive accuracy alone does not make Centaur a reliable participant simulator or an accurate cognitive model, and that generative evaluation must be part of the standard.

What carries the argument

The predictive-versus-generative distinction, operationalized by evaluating models in two modes. In predictive mode, the model receives the true human choice history and is scored by the negative log-likelihood of the next choice. In generative mode, the model receives only its own chosen actions and their feedback, with responses sampled from its softmax at temperature 1. The repetition model—a baseline that repeats its previous choice with fixed probability—is what makes the distinction bite: it predicts human choices reasonably well but, run generatively, repeats the same choice forever and misses the reversal entirely, proving that predictive fit does not certify generative fidelity.

What would settle it

Run Centaur open-loop on the reversal task for 100 fresh seeds and measure the distribution of the trial on which it first switches to the newly rewarded bandit after trial 50. If that switch-latency distribution matches human switch latencies from the underlying reversal-learning data, and if the horizon task reproduces the human horizon effect, the paper's conclusion is wrong.

Watch

Extended reading notes

Core claim

The paper's core discovery is a dissociation: Centaur can achieve strong trial-by-trial predictive performance on tasks from its training set while its open-loop, generative behavior misses qualitative human hallmarks. On the reversal learning task, Centaur shows weak and seed-dependent reversal dynamics, sometimes never switching after the reward reversal; on the horizon task, it does not reproduce the human horizon effect; on the Wisconsin Card Sorting Test, outside its fine-tuning set, it is outperformed by a task-specific model on both predictive and generative measures. Thus, predictive accuracy on its own does not qualify a model as a participant simulator or a cognitive model.

Load-bearing premise

The reversal-learning benchmark treats trajectories generated by a three-parameter RW model ($\alpha=0.5$, $\beta=2.5$, $d=0.5$) as the human reference for generative fidelity; if those synthetic trajectories are not a faithful proxy for human reversal dynamics, that pillar of the conclusion weakens.

Editorial extensions

If this is right

A model can score well on next-choice prediction while failing the generative test that matters for simulation.
Centaur's failure to reproduce the horizon effect and the reversal learning effect indicates it has not learned the decision processes those tasks are designed to expose, despite training on similar task families.
On a task outside its fine-tuning set, Centaur is outperformed by a small task-specific model, suggesting that whatever human-like behavior it does produce does not transfer broadly.
Participant simulators need standardized open-loop generative benchmarks, not just predictive likelihood, before they can support in silico prototyping of experiments.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Extension: if the predictive/generative dissociation generalizes, current next-token prediction scores on cognitive tasks may overstate how aligned LLMs are with humans; adding an open-loop generation version of each benchmark would test this directly.
Extension: the authors' three-task evaluation is a transferable template. Applying the same open-loop protocol to other foundation models trained on human behavior would show whether the failure is specific to Centaur's training objective or endemic to next-token fine-tuning more broadly.
Extension: Centaur's horizon-effect failure suggests a concrete hypothesis—fine-tuning on human choice histories may teach a model to appear human when anchored to human history, but not to carry the internal exploration-exploitation state needed to generate such behavior. A testable extension is prompting Centaur to verbalize its uncertainty before each choice and checking whether that restores t
Extension: because the reversal benchmark uses synthetic trajectories from an RW model as the human reference, a direct replication with real human reversal-learning data would clarify how much of the reported failure is inherent to Centaur and how much is an artifact of the benchmark.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper is worth knowing about: it's the first open-loop generative evaluation of Centaur, and it documents a real predictive-generative dissociation across three cognitive tasks. The methods are transparent—prompts, temperature=1 sampling, per-seed results in the SI—and the horizon and WCST analyses use actual human data. The predictive-generative distinction is well motivated, and the repetition-model toy example in Fig. 1A/B is a nice pedagogical illustration of why prediction can mask generative failures. Credit where due: the negative conclusion is plausible and the paper is a fair, direct test of a claim that has been floating around the LLM-as-participant-simulator literature.

The soft spots are real but not fatal. The flagship reversal-learning 'human' benchmark is not human at all; it's a Rescorla-Wagner simulation with fixed parameters (alpha=0.5, beta=2.5, d=0.5) across 32 seeds. The abstract and conclusion read that result as divergence 'from human data,' which is overclaiming. If the RW model is a poor proxy for human reversal dynamics, the reversal-specific conclusion weakens. However, the horizon and WCST results rely on human data and still show generative divergence, so the central negative result survives even if the reversal column is discounted. Two other minor issues: no inferential statistics (just means and SEs) and only 32 seeds per model. These are more presentation than substance. The conclusion that Centaur 'cannot yet serve as a behavioral AlphaFold' is fine as a title-level claim, but the three-task, one-synthetic-proxy evidence doesn't support a strong systematic claim across all cognitive tasks.

Who is this for? Anyone working on LLM evaluation for cognitive modeling, or on using LLMs as synthetic participants. It's a useful caution for the in-silico-prototyping crowd. It deserves a serious referee, though it would need a revision that either replaces the RW benchmark with human reversal data or explicitly reframes that specific claim as divergence from the domain-specific model rather than from humans. The SI is solid enough that I'd trust the horizon and WCST numbers. It's not a takedown; it's a fair empirical contribution with a overbroad interpretation on one task. I'd accept it for review, and I'd suggest the editor ask for a moderate revision rather than desk reject.

Referee Report

3 major / 4 minor

Summary. The paper evaluates Centaur, an LLM fine-tuned on a large corpus of human behavioral experiments, as a candidate 'participant simulator' in the behavioral sciences. The authors distinguish predictive performance (trial-by-trial NLL given human choice histories) from generative performance (open-loop simulation of entire choice sequences), and argue that the latter is the critical criterion for a simulator. They compare Centaur-7B/80B against Llama-3.1 baselines and domain-specific cognitive models on three tasks: a reversal-learning bandit task, a horizon-dependent bandit task, and the Wisconsin Card Sorting Test. The central empirical claim is a dissociation: Centaur shows competitive or superior predictive accuracy, but its generative behavior deviates from human data, missing qualitative hallmarks such as reversal behavior and horizon effects. The paper concludes that Centaur, despite progress in prediction, is not yet a reliable participant simulator or an accurate model of cognition.

Significance. If the conclusion holds, this is an important and timely cautionary result for the use of LLM-based 'synthetic participants': it would demonstrate that high predictive fit to human choices does not automatically confer generative fidelity, and that standardized generative benchmarks are needed before such models are used for in silico experimental prototyping. The paper has clear strengths: the predictive/generative distinction is well motivated, the evaluation protocols (prompt construction, sampling policy, seed handling) are described transparently, the code and data are openly available, per-seed results are reported in Fig. S1, and two of the three tasks are evaluated against real human datasets. The main limitations—the reversal-learning benchmark uses unvalidated synthetic data, the absence of inferential statistics, and the unexamined dependence on a single sampling policy—mean that the strong version of the conclusion is not yet fully established.

major comments (3)

[Abstract / 'Reversal Learning Task' in Supplementary Information] The abstract and conclusion claim that Centaur's generative behavior 'systematically diverges from human data,' and the main text singles out 'reversals' as a qualitative hallmark Centaur misses. However, for the reversal-learning task the human benchmark is actually synthetic data generated from a three-parameter Rescorla-Wagner model with fixed parameters (alpha=0.5, beta=2.5, d=0.5), as stated in the SI. No evidence is provided that this RW model, or this parameter setting, reproduces the timing, slope, or variability of human reversal behavior. The figure caption is careful ('synthetic data generated from the RW model'), but the abstract and conclusion overstate the human grounding of this specific result. This is load-bearing because the reversal dissociation is the paper's flagship demonstration. Please either validate the RW proxy against human reversal data (or fit its parameters
[Fig. 1D and 1F / 'Generative Performance Evaluation' in Supplementary Information] The conclusion that Centaur 'fails to capture' the horizon effect or exhibits 'substantially more perseveration and set-loss errors' is based on visual inspection of group-mean curves and error bars, with no inferential statistics. For a claim about the absence or attenuation of an effect (e.g., no horizon effect in Fig. 1D), it is important to quantify evidence: report effect sizes and confidence intervals, and for null or near-null effects provide equivalence tests, Bayes factors, or model comparisons on choice data. The simulations used only 32 seeds for the reversal task and similar per-participant timelines elsewhere; a power analysis or per-seed distributions would clarify how stable the qualitative divergences are, especially given the large seed-to-seed variability shown in Fig. S1. This is a central issue because the paper's main claim is an empirical dissociation, not a purely
[Supplementary Information, 'Generative Performance Evaluation' / 'Action Policy'] The generative evaluation treats one specific policy—sampling from the softmax over the constrained answer set at temperature=1 in an open-loop self-feeding regime—as 'Centaur's generative behavior.' But a model's generative capability can be sensitive to decoding parameters (temperature, top-p, constrained token masking, prompt phrasing), and the paper does not test whether the reported divergence is robust across reasonable settings. If the conclusion is meant to be about Centaur as a simulator rather than about one particular sampling policy, the authors should either show robustness to these choices or explicitly restate the conclusion as applying to the temperature-1 constrained policy. This is not a request for exhaustive sweeps, but a minimal sensitivity analysis is needed to support the strong 'systematically diverges' wording.

minor comments (4)

[Supplementary Information, 'Reversal Learning Task'] Typographical/naming inconsistencies: 'Centaur-70B' appears in the reversal generative section but the model is elsewhere described as Centaur-80B; 'Wisconsis' should be 'Wisconsin'; 'extend' should be 'extent'; 'asses' should be 'assess'; 'probaility' should be 'probability'.
[Supplementary Information, 'Reversal Learning Task'] The predictive evaluation of the RW baseline on 'data it had generated itself' is not an independent baseline, since the model is essentially being evaluated on its own training distribution. This is acceptable for illustrating the predictive/generative contrast, but should be acknowledged or supplemented with a cross-validated evaluation on human data.
[Fig. 1 caption] The caption correctly indicates that panel A uses synthetic RW data. Please ensure the main text always consistently uses 'human data' only where human data were actually used, to avoid ambiguity.
[Supplementary Information, 'Reversal Learning Task'] The repetition model's generative behavior is described as 'a random bandit was selected at the start ... and the model repeatedly chose this same arm throughout the timeline with probability p.' It is unclear how the model switches to the other arm when it does not repeat; clarify the exact generative process and whether p is the fitted value or a free parameter varied across seeds.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the evaluation is empirical and self-contained, with one benchmark-support caveat.

full rationale

This is an evaluation paper whose central claim—Centaur's generative behavior diverges from human data—is produced by running the model open-loop on three tasks and comparing its outputs to external human datasets (Wilson et al. for the horizon task, Steinke et al. for the WCST) or to an explicitly labeled RW simulation for the reversal task. The predictive-vs-generative distinction is taken from external references [7,8], not from the authors' own prior work. The self-citations ([3],[5]) motivate the participant-simulator framing but are not load-bearing for any empirical result. The only in-sample evaluation appears in the reversal-task baseline, where the RW model is evaluated on data it generated itself by minimizing NLL; this is an evaluation-quality concern rather than a circular derivation, and it does not affect the paper's main conclusion about Centaur's generative behavior. No equation in the paper is equivalent by construction to its input, and no fitted parameter is renamed as a prediction of Centaur. The abstract's phrasing 'human data' overstates the reversal benchmark, which is RW-simulated, but that is a validity issue, not a circularity issue. Therefore the paper receives a score of 0.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The evaluation relies on standard cognitive-model machinery (softmax choice, Rescorla-Wagner updates, sequential learning attention vectors) and on domain assumptions that LLM next-token probabilities, sampled at temperature=1 over a constrained answer set, stand in for the model's 'behavior'. The paper also assumes RW-synthetic trajectories are a valid human proxy in the reversal task. No new entities are postulated.

free parameters (7)

RW data-generating learning rate alpha = 0.5
Chosen to generate synthetic human trajectories in the reversal learning task; the generative comparison inherits this choice.
RW data-generating inverse temperature beta = 2.5
Chosen for synthetic data generation; affects the sharpness of the reversal behavior used as the benchmark.
RW data-generating initial value d = 0.5
Initial value estimate for both bandits; chosen for synthetic data generation.
Repetition model repeat probability p = not reported
Fit by minimizing NLL on RW-generated data; value not given in the paper.
SL model parameters (r, p, d, f) = r=0.967, p=0.656, d=0.41, f=0.05
Fit globally to 80% of WCST participants; these fitted values drive the domain-specific baseline that outperforms Centaur.
LLM sampling temperature = 1.0
Chosen for generative evaluation to allow behavioral variability between simulated participants; the generative results depend on this choice.
RW fitted parameters (horizon task: alpha, beta, d) = not reported
Fit to 25 participants of the Wilson et al. dataset; values not reported.

assumptions (5)

domain assumption Softmax action selection with Rescorla-Wagner value updates is a valid model of human choice in bandit tasks.
Used as the domain-specific baseline and as the synthetic data generator for the reversal task (Supplementary Information).
domain assumption RW-generated trajectories adequately represent human behavior in the reversal learning task.
The generative comparison in Fig. 1B benchmarks models against RW synthetic data, not human data.
domain assumption The next-token softmax probability over the answer set reflects an LLM's choice behavior.
Both predictive NLL (unmasked human choice) and generative sampling (temperature=1) treat raw logits as behavioral probabilities.
domain assumption The predictive/generative distinction for cognitive models transfers to LLM-based simulators.
Borrowed from Palminteri et al. [8] and Steingroever et al. [7] and applied without modification to LLMs.
standard math Standard model machinery: RW update rule, softmax action selection, sequential learning attention update equations.
Equations in the Supplementary Information are taken as given.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Not Yet AlphaFold for the Mind: Evaluating Centaur as a Synthetic Participant." pith.science (2026). https://pith.science/paper/HF3JPVER

@misc{pith2026250807887,
  author       = {Pith},
  title        = {Pith review of: Not Yet AlphaFold for the Mind: Evaluating Centaur as a Synthetic Participant},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HF3JPVER}},
  note         = {Machine review of arXiv:2508.07887}
}

read the original abstract

Simulators have revolutionized scientific practice across the natural sciences. By generating data that reliably approximate real-world phenomena, they enable scientists to accelerate hypothesis testing and optimize experimental designs. This is perhaps best illustrated by AlphaFold, a Nobel-prize winning simulator in chemistry that predicts protein structures from amino acid sequences, enabling rapid prototyping of molecular interactions, drug targets, and protein functions. In the behavioral sciences, a reliable participant simulator - a system capable of producing human-like behavior across cognitive tasks - would represent a similarly transformative advance. Recently, Binz et al. introduced Centaur, a large language model (LLM) fine-tuned on human data from 160 experiments, proposing its use not only as a model of cognition but also as a participant simulator for "in silico prototyping of experimental studies", e.g., to advance automated cognitive science. Here, we review the core criteria for a participant simulator and assess how well Centaur meets them. Although Centaur demonstrates strong predictive accuracy, its generative behavior - a critical criterion for a participant simulator - systematically diverges from human data. This suggests that, while Centaur is a significant step toward predicting human behavior, it does not yet meet the standards of a reliable participant simulator or an accurate model of cognition.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Small Foundation Models of Human Cognition and Behaviour
cs.AI 2026-08 conditional novelty 6.0 of 10

Tiny cognitively fine-tuned models match a 70B model on familiar experiments, and prompt ablations show they use stimulus and feedback content, not choice-history shortcuts alone.

Reference graph

Works this paper leans on

15 extracted references · 13 canonical work pages · cited by 1 Pith paper

[1]

Nature596(7873), 583–589 (2021)

Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., ˇZ ´ ıdek, A., Potapenko, A.: Highly accurate pro- tein structure prediction with AlphaFold. Nature596(7873), 583–589 (2021). Publisher: Nature Publishing Group

work page 2021
[2]

Physical review letters116(9), 090405 (2016)

Krenn, M., Malik, M., Fickler, R., Lapkiewicz, R., Zeilinger, A.: Automated search for new quantum experiments. Physical review letters116(9), 090405 (2016)

work page 2016
[3]

PsyArXiv

Musslick, S., Strittmatter, Y., Dubova, M.: Closed-loop scientific discovery in the behavioral sciences. PsyArXiv. Preprint (2024). https://doi.org/10.31234/osf.io/ c2ytb.v2

doi:10.31234/osf.io/ 2024
[4]

Nature, 1–8 (2025)

Binz, M., Akata, E., Bethge, M., Br¨ andle, F., Callaway, F., Coda-Forno, J., Dayan, P., Demircan, C., Eckstein, M.K., ´Eltet˝ o, N., et al.: A foundation model to predict and capture human cognition. Nature, 1–8 (2025)

2025
[5]

Proceedings of the National Academy of Sciences122(5), 2401238121 (2025)

Musslick, S., Bartlett, L.K., Chandramouli, S.H., Dubova, M., Gobet, F., Grif- fiths, T.L., Hullman, J., King, R.D., Kutz, J.N., Lucas, C.G.,et al.: Automating the practice of science: Opportunities, challenges, and implications. Proceedings of the National Academy of Sciences122(5), 2401238121 (2025)

work page 2025
[6]

In: Caplin, A., Schotter, A

Gabaix, X., Laibson, D.: The seven properties of good models. In: Caplin, A., Schotter, A. (eds.) The Foundations of Positive and Normative Economics: A Handbook, pp. 292–319. Oxford University Press, Oxford (2008)

work page 2008
[7]

Decision1(3), 161 (2014)

Steingroever, H., Wetzels, R., Wagenmakers, E.-J.: Absolute performance of reinforcement-learning models for the iowa gambling task. Decision1(3), 161 (2014)

work page 2014
[8]

Trends in cognitive sciences21(6), 425–433 (2017)

Palminteri, S., Wyart, V., Koechlin, E.: The importance of falsification in computational cognitive modeling. Trends in cognitive sciences21(6), 425–433 (2017)

work page 2017

Show all 15 references

[9]

Journal of mathematical psychology54(1), 5–13 (2010)

Bishara, A.J., Kruschke, J.K., Stout, J.C., Bechara, A., McCabe, D.P., Buse- meyer, J.R.: Sequential learning models for the wisconsin card sort task: Assessing processes in substance dependent individuals. Journal of mathematical psychology54(1), 5–13 (2010)

2010
[10]

Developmental Cognitive Neuroscience55, 101106 (2022)

Eckstein, M.K., Master, S.L., Dahl, R.E., Wilbrecht, L., Collins, A.G.: Rein- forcement learning and bayesian inference provide complementary models for the unique advantage of adolescents in stochastic reversal. Developmental Cognitive Neuroscience55, 101106 (2022)

2022
[11]

Journal 12 of experimental psychology: General143(6), 2074 (2014)

Wilson, R.C., Geana, A., White, J.M., Ludvig, E.A., Cohen, J.D.: Humans use directed and random exploration to solve the explore–exploit dilemma. Journal 12 of experimental psychology: General143(6), 2074 (2014)
[12]

Scientific reports10(1), 15464 (2020)

Steinke, A., Lange, F., Kopp, B.: Parallel model-based and model-free rein- forcement learning for card sorting performance. Scientific reports10(1), 15464 (2020)

2020
[13]

Neuroscience345, 12– 26 (2017)

Izquierdo, A., Brigman, J.L., Radke, A.K., Rudebeck, P.H., Holmes, A.: The neural basis of reversal learning: An updated perspective. Neuroscience345, 12– 26 (2017)

2017
[14]

Journal of experimental psychology38(4), 404 (1948)

Grant, D.A., Berg, E.: A behavioral analysis of degree of reinforcement and ease of shifting to new responses in a weigl-type card-sorting problem. Journal of experimental psychology38(4), 404 (1948)

1948
[15]

Brain and cognition 71(3), 437–451 (2009) 13

Nyhus, E., Barcel´ o, F.: The wisconsin card sorting test and the cognitive assess- ment of prefrontal executive functions: a critical update. Brain and cognition 71(3), 437–451 (2009) 13

2009

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Nature596(7873), 583–589 (2021)

Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., ˇZ ´ ıdek, A., Potapenko, A.: Highly accurate pro- tein structure prediction with AlphaFold. Nature596(7873), 583–589 (2021). Publisher: Nature Publishing Group

work page 2021

[2] [2]

Physical review letters116(9), 090405 (2016)

Krenn, M., Malik, M., Fickler, R., Lapkiewicz, R., Zeilinger, A.: Automated search for new quantum experiments. Physical review letters116(9), 090405 (2016)

work page 2016

[3] [3]

PsyArXiv

Musslick, S., Strittmatter, Y., Dubova, M.: Closed-loop scientific discovery in the behavioral sciences. PsyArXiv. Preprint (2024). https://doi.org/10.31234/osf.io/ c2ytb.v2

doi:10.31234/osf.io/ 2024

[4] [4]

Nature, 1–8 (2025)

Binz, M., Akata, E., Bethge, M., Br¨ andle, F., Callaway, F., Coda-Forno, J., Dayan, P., Demircan, C., Eckstein, M.K., ´Eltet˝ o, N., et al.: A foundation model to predict and capture human cognition. Nature, 1–8 (2025)

2025

[5] [5]

Proceedings of the National Academy of Sciences122(5), 2401238121 (2025)

Musslick, S., Bartlett, L.K., Chandramouli, S.H., Dubova, M., Gobet, F., Grif- fiths, T.L., Hullman, J., King, R.D., Kutz, J.N., Lucas, C.G.,et al.: Automating the practice of science: Opportunities, challenges, and implications. Proceedings of the National Academy of Sciences122(5), 2401238121 (2025)

work page 2025

[6] [6]

In: Caplin, A., Schotter, A

Gabaix, X., Laibson, D.: The seven properties of good models. In: Caplin, A., Schotter, A. (eds.) The Foundations of Positive and Normative Economics: A Handbook, pp. 292–319. Oxford University Press, Oxford (2008)

work page 2008

[7] [7]

Decision1(3), 161 (2014)

Steingroever, H., Wetzels, R., Wagenmakers, E.-J.: Absolute performance of reinforcement-learning models for the iowa gambling task. Decision1(3), 161 (2014)

work page 2014

[8] [8]

Trends in cognitive sciences21(6), 425–433 (2017)

Palminteri, S., Wyart, V., Koechlin, E.: The importance of falsification in computational cognitive modeling. Trends in cognitive sciences21(6), 425–433 (2017)

work page 2017