A habit and working memory model as an alternative account of human reward-based learning

Reinforcement learning (RL) algorithms have had tremendous success accounting for reward-based learning across species, including instrumental learning in contextual bandit tasks, and they capture variance in brain signals. However, reward-based learning in humans recruits multiple processes, including memory and choice perseveration; their contributions can easily be mistakenly attributed to RL computations. Here I investigate how much of reward-based learning behaviour is supported by RL computations in a context where other processes can be factored out. Reanalysis and computational modelling of 7 datasets (n = 594) in diverse samples show that in this instrumental context, reward-based learning is best explained by a combination of a fast working-memory-based process and a slower habit-like associative process, neither of which can be interpreted as a standard RL-like algorithm on its own. My results raise important questions for the interpretation of RL algorithms as capturing a meaningful process across brain and behaviour.