Jian Li

This paper introduces DynamicRL, a method to estimate how humans adapt their reinforcement learning strategies over time, showing parameter updates resemble policy gradient ascent.

This study compares parameter estimation methods in reinforcement learning models, finding that deep learning outperforms traditional optimization in reliability and generalizability.

This paper demonstrates that complex synapses enhance memory capacity for face familiarity detection, outperforming simple synapses in real-world recognition tasks.