Matthew Botvinick

This paper introduces DataDIVER, an automated method for discovering computational models from data, which reveals novel insights into human and animal learning.

This study shows how rapid learning in mice involves structured changes in motor cortex connectivity, rerouting preparatory activity to control a conditioned neuron.

This paper develops a detailed biomechanical model of fruit flies for simulating locomotion using reinforcement learning.

This paper introduces DiscoGen, a neural network-based method for discovering Gene Regulatory Networks that outperforms state-of-the-art causal discovery methods.

This paper learns cognitive models from behavioral data using disentangled RNNs, recovering interpretable models from synthetic and rat behavioral data.

This study shows that two BCI learning phenomena can be explained by gradient-based optimization, without requiring special mechanisms.

This study investigates whether value representations in the rodent orbitofrontal cortex drive choice or support learning, finding that they instead signal expected reward information to a learning process elsewhere in the brain.

The paper provides neural evidence for the successor representation in choice evaluation by linking behavioral errors to predictive representations in sensory cortex.

The paper shows that pre-trial firing rate patterns in human dlPFC predict conflict resolution in cognitive tasks, with distinct but related patterns for different conflict types.

The paper shows that visual understanding involves relational and compositional representations in the prefrontal-hippocampal circuit, with replay sequences forming hypotheses about scene configurations.

This paper introduces DGNs, a biologically plausible learning rule that outperforms backpropagation in efficiency and forgetting resistance, inspired by cerebellar structure and function.

The paper shows that conflict in the dorsal anterior cingulate cortex (dACC) amplifies task-relevant information across neurons, rather than relying on specific conflict-encoding neurons or coactivation of action plans.

This study investigates prediction error signals in medial prefrontal cortex (mPFC) during goal and subgoal tasks, showing mPFC signals subgoal-related errors independently and context-dependent goal-related errors.

The paper extends a unified account of reinforcement learning to include episodic learning by introducing a meta-learner that coordinates different learning algorithms.

This paper investigates how the brain integrates evidence for different perceptual attributes and responses during decision-making, finding distinct roles for ventromedial prefrontal cortex and dorsal anterior cingulate cortex.

The paper proposes a new theory of reward-based learning where the prefrontal cortex functions as a meta-reinforcement learning system trained by the dopamine system.

This paper proposes that the hippocampus uses a predictive representation for spatial navigation, challenging the traditional cognitive map metaphor and suggesting grid cells support hierarchical planning.

This paper proposes a framework linking model-based reinforcement learning to model-free mechanisms using successor representations and introduces algorithms that mitigate their limitations.

This study investigates the successor representation (SR) in human reinforcement learning, showing it balances flexibility and efficiency by being insensitive to task sequence changes but responsive to reward changes.

This study provides causal evidence linking the dorsal hippocampus to planning behavior using a novel task and muscimol inactivation in rodents.

The paper presents a new analysis technique for multi-step decision tasks that reduces confounding issues and provides a theory-neutral view of behavior.