Self-distillation token rewards measure input-response-feedback pointwise mutual information, and CREDIT extracts the input-specific component with contrastive baselines to improve LLM reasoning performance.
Policy invariance under reward transforma- tions: Theory and application to reward shaping
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
EvoNav automates the design of reward functions for RL robot navigation by evolving LLM proposals through a three-stage cheap-to-expensive evaluation process and claims better policies than hand-crafted or prior automated rewards.
SBC generates virtual environments via state blocking to expose agents to diverse suboptimal partner policies, yielding superior zero-shot coordination performance including with humans.
citing papers explorer
-
From Generic Correlation to Input-Specific Credit in On-Policy Self Distillation
Self-distillation token rewards measure input-response-feedback pointwise mutual information, and CREDIT extracts the input-specific component with contrastive baselines to improve LLM reasoning performance.
-
EvoNav: Evolutionary Reward Function Design for Robot Navigation with Large Language Models
EvoNav automates the design of reward functions for RL robot navigation by evolving LLM proposals through a three-stage cheap-to-expensive evaluation process and claims better policies than hand-crafted or prior automated rewards.
-
Shaping Zero-Shot Coordination via State Blocking
SBC generates virtual environments via state blocking to expose agents to diverse suboptimal partner policies, yielding superior zero-shot coordination performance including with humans.