Gate-Rheology: Inertia of Cognitive Control Explains Meta-Rigidity in Sequential Decision Making and Reversal Learning

Алексей Леонидович Снигиров

ПРЕПРИНТ

Эта статья является препринтом и не была отрецензирована.
О результатах, изложенных в препринтах, не следует сообщать в СМИ как о проверенной информации.

Gate-Rheology: Inertia of Cognitive Control Explains Meta-Rigidity in Sequential Decision Making and Reversal Learning

А. Л. Снигиров

2026-03-30

https://doi.org/10.24108/preprints-3114801

Background. Cognitive flexibility—the ability to adapt behavior to changing environmental contingencies—is paradoxically constrained by systematic forms of rigidity. Organisms persist with outdated strategies long after they cease to be adaptive, yet traditional computational models assume instantaneous arbitration between habitual and goal-directed control. Methods. We introduce Gate-Rheology, a mechanistic framework in which arbitration between computational modes possesses intrinsic inertia, formalized as a viscosity parameter V_G that resists mode switching. We validate this framework through two canonical paradigms: Two-Step Task (Stage 2A) and Block-Reversal Learning (Stage 2B), using ablation studies to dissociate control-mode inertia (V_G) from action-level perseveration (V_p). Results. Across 30 independent seeds, RheologicalAgent reproduced canonical MB/MF signatures (reward × transition interaction: coef = 0.312 ± 0.118, p = 0.008) while exhibiting heavy-tailed mode-switch latencies after changepoints (median = 35 trials, max = 699). Ablation of V_G eliminated switching latency (log-rank χ² = 42.1, p = 8.48 × 10⁻¹¹), while ablation of V_p selectively reduced perseverative errors (median = 3 vs. 8 for Full; p = 5.85 × 10⁻⁵). In the Reversal task, NoVp robustly reduced perseverative errors, while the planned latency comparison did not reach significance; survival analysis suggested only a weak secondary trend. The same agent with identical parameters succeeded in both tasks, demonstrating cross-paradigm generalization. Conclusion. Gate-Rheology provides a mechanistic account of cognitive rigidity grounded in dynamic control-mode arbitration rather than static value representations. The functional dissociation between control-mode inertia and action-level perseveration (between V_G and V_p) suggests that meta-cognitive inertia and behavioral stickiness are computationally distinct mechanisms requiring separate experimental manipulation. Predictions for rodent reversal learning and clinical phenotypes are discussed.

Ссылка для цитирования:

Снигиров А. Л. 2026. Gate-Rheology: Inertia of Cognitive Control Explains Meta-Rigidity in Sequential Decision Making and Reversal Learning. PREPRINTS.RU. https://doi.org/10.24108/preprints-3114801

Список литературы

1. Akam, T., Costa, R., & Dayan, P. (2015). Simple plans or sophisticated habits? State, transition and learning interactions in the two-step task. PLOS Computational Biology, 11(12), e1004648. doi:10.1371/journal.pcbi.1004648

2. Brands, A. M., Mathar, D., & Peters, J. (2025). Signatures of perseveration and heuristic-based directed exploration in two-step sequential decision task behaviour. Computational Psychiatry, 9(1), 39–62. https://doi.org/10.5334/cpsy.101

3. Daw, N. D., Gershman, S. J., Seymour, B., Dayan, P., & Dolan, R. J. (2011). Model-based influences on humans’ choices and striatal prediction errors. Neuron, 69(6), 1204–1215. https://doi.org/10.1016/j.neuron.2011.02.027

4. den Ouden, H. E. M., Daw, N. D., Fernandez, G., Elshout, M., Rijpkema, M., Hoogman, M., Franke, B., & Cools, R. (2013). Dissociable effects of dopamine and serotonin on reversal learning. Neuron, 80(4), 1090–1100. https://doi.org/10.1016/j.neuron.2013.08.030

5. Dias, R., Robbins, T. W., & Roberts, A. C. (1996). Primate analogue of the Wisconsin Card Sorting Test: Effects of excitotoxic lesions of the prefrontal cortex in the marmoset. Behavioral Neuroscience, 110(5), 872–886. https://doi.org/10.1037/0735-7044.110.5.872

6. Dolan, R. J., & Dayan, P. (2013). Goals and habits in the brain. Neuron, 80(2), 312–325. https://doi.org/10.1016/j.neuron.2013.09.007

7. Feher da Silva, C., & Hare, T. A. (2018). A note on the analysis of two-stage task results: How changes in task structure affect what model-free and model-based strategies predict about the effects of reward and transition on the stay probability. PLOS ONE, 13(4), e0195328. https://doi.org/10.1371/journal.pone.0195328

8. Hauser, T. U., Moutoussis, M., Iannaccone, R., Brem, S., Walitza, S., Drechsler, R., Dayan, P., & Dolan, R. J. (2017). Increased fronto-striatal reward prediction errors moderate decision making in obsessive-compulsive disorder. Psychological Medicine, 47(7), 1246–1258. https://doi.org/10.1017/S003329171600356X

9. Hasz, B. M., & Redish, A. D. (2018). Deliberation and procedural automation on a two-step task for rats. Frontiers in Integrative Neuroscience, 12, 30. https://doi.org/10.3389/fnint.2018.00030

10. Izquierdo, A., Brigman, J. L., Radke, A. K., Rudebeck, P. H., & Holmes, A. (2017). The neural basis of reversal learning: An updated perspective. Neuroscience, 345, 12–26. https://doi.org/10.1016/j.neuroscience.2016.03.021

11. Kanen, J. W., Ersche, K. D., Fineberg, N. A., Robbins, T. W., & Cardinal, R. N. (2019). Computational modelling reveals contrasting effects on reinforcement learning and cognitive flexibility in stimulant use disorder and obsessive-compulsive disorder. Psychopharmacology, 236(8), 2337–2358. https://doi.org/10.1007/s00213-019-05325-w

12. Keramati, M., Dezfouli, A., & Piray, P. (2011). Speed/accuracy trade-off between the habitual and the goal-directed processes. PLOS Computational Biology, 7(5), e1002055. https://doi.org/10.1371/journal.pcbi.1002055

13. Le, N. M., Yildirim, M., Wang, Y., Sugihara, H., Jazayeri, M., & Sur, M. (2023). Mixtures of strategies underlie rodent behavior during reversal learning. PLOS Computational Biology, 19(9), e1011430. https://doi.org/10.1371/journal.pcbi.1011430

14. Lee, S. W., Shimojo, S., & O’Doherty, J. P. (2014). Neural computations underlying arbitration between model-based and model-free learning. Neuron, 81(3), 687–699. https://doi.org/10.1016/j.neuron.2013.11.028

15. Miller, K. J., Botvinick, M. M., & Brody, C. D. (2017). Dorsal hippocampus contributes to model-based planning. Nature Neuroscience, 20(9), 1242–1247. https://doi.org/10.1038/nn.4611

16. Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7(3), 134–140. https://doi.org/10.1016/S1364-6613(03)00028-7

17. Otto, A. R., Raio, C. M., Chiang, A., Phelps, E. A., & Daw, N. D. (2013). Working-memory capacity protects model-based learning from stress. Proceedings of the National Academy of Sciences, 110(52), 20941–20946. https://doi.org/10.1073/pnas.1312011110

18. Redish, A. D. (2016). Vicarious trial and error. Nature Reviews Neuroscience, 17(3), 147–159. https://doi.org/10.1038/nrn.2015.30

19. Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013). The expected value of control: An integrative theory of anterior cingulate cortex function. Neuron, 79(2), 217–240. https://doi.org/10.1016/j.neuron.2013.07.007

20. Snigirov, A. L. (2025). Principia Cognitia: An axiomatic framework for the new science of mind [Preprint]. PhilArchive. https://philarchive.org/rec/SNIPCA

21. Wilson, R. C., & Collins, A. G. E. (2019). Ten simple rules for the computational modeling of behavioral data. eLife, 8, e49547. https://doi.org/10.7554/eLife.49547

22. Wilson, R. C., Geana, A., White, J. M., Ludvig, E. A., & Cohen, J. D. (2014). Humans use directed exploration to reduce uncertainty when choosing between options. Proceedings of the 36th Annual Conference of the Cognitive Science Society, 3632–3637.

ПРЕПРИНТ

ИНФОРМАЦИЯ

Использование куки-файлов