Leo Chi U Seak

The paper demonstrates that humans reuse learned subtask representations during planning, showing behavioral, computational, and neural evidence of subtask reactivation and reuse.

The paper investigates neural dynamics during multi-step planning using MEG, showing evolving value encoding and state reactivation linked to improved performance and learning.

The paper shows that dopamine neurons in monkeys encode subjective reward value by integrating utility and subjectively weighted probability, aligning with observed choice behavior.

This study compares risky choices between humans and monkeys, finding comparable decision-making patterns despite differences in subjective values.

The paper proposes and computationally validates a method to mimic artificial neural networks using genetic circuits, demonstrating functionality in regression and classification analyses.

The paper demonstrates that neural signals in reward-related brain regions follow choice indifference curves, indicating integration of multiple reward components into scalar signals.

The paper characterizes Azotobacter vinelandii for synthetic biology and develops two BioBrick kits for regulated protein expression and magnetosome conjugation.