Nathaniel D. Daw

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 that meta-learning in rats involves changes in prefrontal cortical dynamics that enable rule-based reward learning.

The paper replicates a previous finding that cognitive effort is influenced by reward rate but shows that recent performance history better explains effort allocation, suggesting a reinterpretation of prior results.

The paper investigates how neural replay relates to latent cause inference and supports fast generalization in a value-based task with shared rewards.

This study shows that the brain uses shared neural subspaces to flexibly perform and combine multiple tasks by reusing representations across different task-related information.

This paper presents a computational theory of infant gaze based on rational information gathering, showing that gaze preference is nonmonotonic with exposure due to the interplay of stimulus informativeness and future utility.

The paper investigates how humans learn successor representations using trial-by-trial reaction times in graph sequence learning tasks, finding that eligibility trace mechanisms play a predominant role.

This paper introduces a model for efficient model learning and planning using compositional predictive maps, showing it simulates response fields in the medial entorhinal cortex.

The paper shows error-driven drift in hippocampal representations enables flexible re-learning by allowing new associations to form while maintaining previous ones in synaptic weights.

The paper shows that the brain uses proactive and reactive memory strategies for decision-making, with proactive being more efficient but reactive used when options have more memory associations.

LLM-driven evolutionary program synthesis (CogFunSearch) discovers interpretable Python cognitive models that predict bandit-task choices of humans, rats and fruit flies better than leading hand-designed models.

The paper investigates how humans balance detailed and temporally abstract world models during learning, finding individual variation in strategy use and dynamic adaptation based on environmental predictability.

The paper shows that pre-existing visual responses in dopamine terminals predict individual learning trajectories in mice during a visual task.

This study compares task-switching dynamics between artificial neural networks and human EEG data using state-space modeling, revealing a shared strategy of converging to a neutral task state.

The paper shows that rat hippocampus generates flexible representations of alternative spatial locations during navigation, modulated by changing cognitive demands.

The paper introduces a feature-specific reward prediction error model to explain heterogeneity in dopamine neuron responses in the VTA and shows it aligns with neural recordings.

The paper introduces a magnetic voluntary head-fixation system for longitudinal imaging of hippocampal neurons in transgenic rats.

This paper introduces a theory of hippocampal replay that explains how replay builds a cognitive map for navigation, reconciling conflicting hypotheses about its function and predicting new experimental outcomes.

The paper introduces a dynamic model to analyze strategy shifts in reward learning, showing a progression from model-based exploration to exploitation and reduced engagement in rats.

This study investigates how the brain learns and uses attentional templates in decision-making tasks, showing that templates are represented in prefrontal and parietal cortex and transform stimuli into value representations.

The paper develops an Effort Foraging Task to measure cognitive and physical effort costs in humans and finds they are positively correlated.

The paper shows dopamine signals in rats update value through two learning processes during foraging in a complex environment.

The paper combines artificial neural networks with classic cognitive models to create interpretable hybrid models that reveal human learning mechanisms similar to reinforcement learning.

The paper investigates how latent cause inference can lead to instrumental insensitivity to reinforcer devaluation without requiring model-free mechanisms.

The paper investigates how uncertainty affects the balance between incremental learning and episodic memory in decision-making, finding that people rely more on episodic memory when incremental learning is uncertain.

The paper investigates how monkeys switch between learned rules and continuously update them during a perceptual categorization task, revealing distinct computational processes for rule switching and rule learning.

The paper investigates how confidence in others' decisions is computed by combining estimates of ability and decision difficulty, supported by specific brain systems.

This study shows that rat anterior cingulate cortex activity reflects dynamic encoding of decision variables in a patch foraging task, but these signals are not directly involved in decision-making.

The paper develops a Bayesian computational model to predict individual differences in non-reinforced preference changes during Cue-Approach Training.

The paper tests how reward prediction affects memory, finding that high-reward items are better remembered when mixed with low-reward items, but forgetting rates are not influenced by reward.