Yali Du

This article discusses how causal concepts enhance AI's ability to generalize to new data.

This study presents a method for allocating rewards in unknown environments.

NLRL combines reinforcement learning with natural language for improved decision-making.

AgA offers a way to align personal and collective goals in mixed-motive cooperation.

New method shows promise in solving complex combinatorial challenges.

A new method enables agents to follow safety rules in plain language.

Study on improving discussion strategies for AI in One Night Ultimate Werewolf.

A new method enhances how agents predict future rewards and explain decisions.

A new method improves translation across multiple programming languages.

Manipulating AI behavior poses serious risks in advanced systems.