Mian Zhang

A novel two-axis framework unifies the study of multimodal agent memory, revealing distinct tiers of memory adaptation from static reuse to self-evolving policies.

Exploration bias in RL leads models to favor easy instructions, but a novel two-stage framework can unlock their potential for more challenging tasks.

Reward hacking can be mitigated with a simple one-line fix that improves out-of-distribution performance while keeping training robust.