ARF-RLHF: Adaptive Reward-Following for RLHF through Emotion-Driven Self-Supervision and Trace-Biased Dynamic Optimization.

Yuxuan Zhang

Browse the full ACL paper archive.