Cost-Benefit Arbitration Between Multiple Reinforcement-Learning Systems
Open accessThe DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Abstract
Human behavior is sometimes determined by habit and other times by goal-directed planning. Modern reinforcement-learning theories formalize this distinction as a competition between a computationally cheap but inaccurate model-free system that gives rise to habits and a computationally expensive but accurate model-based system that implements planning. It is unclear, however, how people choose to allocate control between these systems. Here, we propose that arbitration occurs by comparing each system’s task-specific costs and benefits. To investigate this proposal, we conducted two experiments showing that people increase model-based control when it achieves greater accuracy than model-free control, and especially when the rewards of accurate performance are amplified. In contrast, they are insensitive to reward amplification when model-based and model-free control yield equivalent accuracy. This suggests that humans adaptively balance habitual and planned action through on-line cost-benefit analysis.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Reinforcement Learning: An Introduction
Sutton, R. S., & Barto, A. G. (1998) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.1998.712192 · Cited by 27,225
A perspective on judgment and choice: mapping bounded rationality.
Kahneman, D. (2003) · American Psychologist · doi:10.1037/0003-066x.58.9.697 · Cited by 5,540
Daw, N. D., Niv, Y., & Dayan, P. (2005) · Nature Neuroscience · doi:10.1038/nn1560 · Cited by 2,533
The expected value of control: an integrative theory of anterior cingulate cortex function
Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013) · Neuron · doi:10.1016/j.neuron.2013.07.007 · Cited by 2,321
Adaptive strategy selection in decision making.
Payne, J. W., Bettman, J. R., & Johnson, E. J. (1988) · Journal of Experimental Psychology: Learning, Memory, and Cognition · doi:10.1037/0278-7393.14.3.534 · Cited by 1,955
An opportunity cost model of subjective effort and task performance
Kurzban, R., Duckworth, A., Kable, J. W., & Myers, J. (2013) · Behavioral and Brain Sciences · doi:10.1017/s0140525x12003196 · Cited by 1,381
Actions and habits: The development of behavioural autonomy
Dickinson, A. (1985) · Philosophical Transactions of the Royal Society of London. B, Biological Sciences · doi:10.1098/rstb.1985.0010 · Cited by 1,217
Gläscher, J., Daw, N., Dayan, P., & O'Doherty, J. P. (2010) · Neuron · doi:10.1016/j.neuron.2010.04.016 · Cited by 1,166
Decision making and the avoidance of cognitive demand
Kool, W., McGuire, J. T., Rosen, Z. B., & Botvinick, M. M. (2010) · Journal of Experimental Psychology: General · doi:10.1037/a0020198 · Cited by 1,141
Self-citationAnimal intelligence; experimental studies
Thorndike, E. L. (1911) · doi:10.5962/bhl.title.55072 · Cited by 1,100
Cited by
Explanations Can Reduce Overreliance on AI Systems During Decision-Making
Vasconcelos, H., Jörke, M., Grunde-McLaughlin, M., Gerstenberg, T., Bernstein, M. S., & Krishna, R. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3579605 · Cited by 402
Kool, W., & Botvinick, M. (2018) · Nature Human Behaviour · doi:10.1038/s41562-018-0401-9 · Cited by 288
Self-citationCushman, F. (2020) · Behavioral and Brain Sciences · doi:10.1017/s0140525x19001730 · Cited by 260
Self-citationTen-month-old infants infer the value of goals from the costs of actions
Liu, S., Ullman, T. D., Tenenbaum, J. B., & Spelke, E. S. (2017) · Science · doi:10.1126/science.aag2132 · Cited by 230
Humans primarily use model-based inference in the two-stage task
Feher da Silva, C., & Hare, T. A. (2020) · Nature Human Behaviour · doi:10.1038/s41562-020-0905-y · Cited by 153
The opportunity cost of time modulates cognitive effort
Otto, A. R., & Daw, N. D. (2019) · Neuropsychologia · doi:10.1016/j.neuropsychologia.2018.05.006 · Cited by 138
Reasoning supports utilitarian resolutions to moral dilemmas across diverse measures.
Patil, I., Zucchelli, M. M., Kool, W., Campbell, S., Fornasier, F., Calò, M., Silani, G., Cikara, M., & Cushman, F. (2021) · Journal of Personality and Social Psychology · doi:10.1037/pspp0000281 · Cited by 133
Self-citationRational use of cognitive resources in human planning
Callaway, F., van Opheusden, B., Gul, S., Das, P., Krueger, P. M., Griffiths, T. L., & Lieder, F. (2022) · Nature Human Behaviour · doi:10.1038/s41562-022-01332-8 · Cited by 121
“For example, people might limit the depth of their search 4–7 , ‘prune’ away initially unpromising courses of action 1,2 or avoid planning altogether by relying on habit or ‘memoization’ 1,15 .”
BackgroundIncentives Boost Model-Based Control Across a Range of Severity on Several Psychiatric Constructs
Patzelt, E. H., Kool, W., Millner, A. J., & Gershman, S. J. (2019) · Biological Psychiatry · doi:10.1016/j.biopsych.2018.06.018 · Cited by 94
Self-citationRecent developments in the habit hypothesis of OCD and compulsive disorders
Gillan, C. M. (2021) · Current Topics in Behavioral Neurosciences · doi:10.1007/7854_2020_199 · Cited by 27
Time spent thinking in online chess reflects the value of computation
Russek, E., Acosta-Kane, D., van Opheusden, B., Mattar, M. G., & Griffiths, T. (2022) · doi:10.31234/osf.io/8j9zx · Cited by 18
Relational rule discovery in complex discrimination learning.
Don, H. J., Goldwater, M. B., Greenaway, J. K., Hutchings, R., & Livesey, E. J. (2020) · Journal of Experimental Psychology: Learning, Memory, and Cognition · doi:10.1037/xlm0000848 · Cited by 12
Colas, J. T., O’Doherty, J. P., & Grafton, S. T. (2024) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1011950 · Cited by 6
Luna, R., Vadillo, M. A., & Luque, D. (2023) · Cortex · doi:10.1016/j.cortex.2023.06.009 · Cited by 3
Zhang, S., Yi, X., Li, S., Hong, C., Chen, G., Liu, J., Wang, X., Hu, Y., Wang, Y., & Li, H. (2025) · arXiv · doi:10.48550/arxiv.2502.06179 · Cited by 1
Intention–Outcome Trade‐Off in Moral Character Learning
Huang, G., Xu, Y., Derrington, E., Dreher, J., & Qu, C. (2026) · Annals of the New York Academy of Sciences · doi:10.1111/nyas.70233
Metadata
- Authors
- Kool, W., Gershman, S. J., & Cushman, F. A.
- Venue
- Psychological Science
- Publisher
- SAGE Publications
- Published
- 2017-07-21
- Volume
- 28
- Issue
- 9
- Pages
- 1321-1333
- ISSN
- 0956-7976 (ISSN-L)
- 1467-9280
- OpenAlex
- W2736983092
- Cited by
- 237
- References
- 20
- FWCI
- 11.17
- Type
- journal article
- Language
- en
- Open access
- green
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.
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.