5 papers
Be My Tutor: On-Policy Co-Distillation for Mutual LLM Improvement via Peer Feedback
Woohyeon Byeon, Jiwon Jeon, Jeonghye Kim +1
We study multi-domain LLM training in which two models, each stronger in a different domain, co-evolve by tutoring each other through on-policy feedback. Unlike one-way distillatio…
Adaptive Action Chunking via Multi-Chunk Q Value Estimation
Yongjae Shin, Jongseong Chae, Seongmin Kim +2
Action chunking emerged as a pivotal technique in imitation learning, enabling policies to predict cohesive action sequences rather than single actions. Recently, this approach has…
Align While Search: Belief-Guided Exploratory Inference for World-Grounded Embodied Agents
Seohui Bae, Jeonghye Kim, Youngchul Sung +1
In this paper, we propose a test-time adaptive agent that performs exploratory inference through posterior-guided belief refinement without relying on gradient-based updates or add…
ReflAct: World-Grounded Decision Making in LLM Agents via Goal-State Reflection
Jeonghye Kim, Sojeong Rhee, Minbeom Kim +4
Recent advances in LLM agents have largely built on reasoning backbones like ReAct, which interleave thought and action in complex environments. However, ReAct often produces ungro…
Penalizing Infeasible Actions and Reward Scaling in Reinforcement Learning with Offline Data
Jeonghye Kim, Yongjae Shin, Whiyoung Jung +5
Reinforcement learning with offline data suffers from Q-value extrapolation errors. To address this issue, we first demonstrate that linear extrapolation of the Q-function beyond t…