Sen Cheng

Human cerebellum and ventral tegmental area interact during extinction of learned fear

This study shows that the cerebellum and ventral tegmental area interact during fear extinction learning, supporting a role for prediction-error processing in this process.

Competition Between Memory Updating and Differentiation Emerges from Intrinsic Network Dynamics

The paper presents a computational model based on a modified Hopfield network to explain how prediction errors drive competition between memory updating and differentiation in recognition tasks.

Accounting for sensitivity of latent learning to behavioral statistics with successor representations

This study shows that deep successor representations can explain how targeted pre-exposure improves latent learning in spatial navigation tasks by aligning spatial features with reward locations.

The first step is not always the hardest: A change-point analysis of predictive learning

This paper uses change-point analysis to show that reversal learning is more difficult than initial learning, as evidenced by later change points and hippocampus-dependent replay of previous experiences.

Unifying spatial and episodic representations in the hippocampus through flexible memory use

This paper proposes that the hippocampus's primary function is storing and retrieving episodic memories, with spatial representations emerging as a byproduct, using a computational model to demonstrate this.

Place cell activity and behaviour during task acquisition and extinction learning predict renewal outcome

The paper examines how place cell activity during learning and extinction predicts renewal outcomes in rats, showing that global remapping correlates with successful renewal.

Divide and conquer: How avian 'prefrontal' and hippocampal neurons process extinction learning in complementary ways

This study shows that pigeon hippocampus and prefrontal NCL process extinction learning differently, with HPC encoding context phases and NCL enabling decision-making and meta-learning.

Global remapping emerges as the mechanism for renewal of context-dependent behavior in a reinforcement learning model

This study uses a reinforcement learning model to show that global remapping of spatial representations underlies context-dependent behavior renewal.

Cerebellar transcranial alternating current stimulation in the theta band facilitates extinction of learned fear responses

This study uses 6 Hz cerebellar transcranial alternating current stimulation to enhance fear extinction learning by modulating cerebellar-cortical interactions in healthy participants.

Dopaminergic drugs modulate fear extinction related processes in humans, but effects are mild

This study examines the effects of dopaminergic drugs on fear extinction learning in humans, finding small effects primarily in the bromocriptine group.

The cost of behavioral flexibility: reversal learning driven by a spiking neural network

This paper models reversal learning in spatial navigation using spiking neural networks and identifies neural mechanisms that enable behavioral flexibility with associated costs.

Competition and integration of visual and goal vector signals for spatial navigation

The paper presents a computational model of visual and goal-vector signal integration in spatial navigation using reinforcement learning, showing that independent use and integration of signals are crucial for navigation in uncertain environments.

Working memory performance is tied to stimulus complexity

The paper investigates how stimulus complexity affects working memory performance by analyzing neural coding patterns in avian prefrontal cortex.

Phase-precession-like Effect in the Anterior Insula Cortex during Reward Expectancy

The paper identifies a phase-precession-like effect in the anterior insula cortex during reward expectancy, linking oscillatory coordination to improved behavioral performance.

Learning to predict future locations with internally generated theta sequences

The paper presents a computational model that generates theta sequences to predict future locations based on internal dynamics and behavioral patterns.

Navigation and the Efficiency of Spatial Coding: Insights from Closed-Loop Simulations

This study uses closed-loop simulations to show that efficient spatial coding, particularly through place cell parameters and Fisher information, is critical for navigation performance.

A Model of Hippocampal Replay Driven by Experience and Environmental Structure Facilitates Spatial Learning

This paper introduces a stochastic replay mechanism for reinforcement learning agents that improves spatial learning performance by prioritizing experiences based on strength, similarity, and inhibition of return.

CoBeL-RL: A neuroscience-oriented simulation framework for complex behavior and learning

CoBeL-RL is a simulation framework for modeling complex behavior and learning in neuroscience, offering virtual environments and RL algorithms for efficient simulation and analysis.

Navigation task and action space drive the emergence of egocentric and allocentric spatial representations

This study shows that navigation tasks influence the type of spatial representations (egocentric or allocentric) that emerge in a deep reinforcement learning model.

The functional role of episodic memory in spatial learning

This study investigates how episodic memory facilitates spatial learning through retrieval and replay modes, finding that replay learning achieves better long-term performance in complex tasks with limited trials.

What happened when? Brain and behavioral responses to modified structure and content in episodic cueing

The paper investigates how different types of episodic prediction errors affect brain activity and memory performance using modified episodic cues and fMRI.

Behavior-dependent spatial maps enable efficient theta phase coding

The paper introduces a new coding scheme for hippocampal theta phase that accounts for all experimental observations by combining features of spatial and temporal sweeps.

Context-dependent extinction learning emerging from raw sensory inputs: A reinforcement learning approach

This study develops a reinforcement learning model that learns context-dependent extinction behavior from raw visual inputs, showing that context dependence can emerge without explicit context representation.