Showing 2026Show all
2 papers · 1 filter
cs.LG2026
Transfer-Stratified On-Policy Distillation for RL-Improved Reasoning Teachers
Xiaoyu Chen, Bo Shao, Tiangang Zhu +5
Reinforcement learning can substantially improve a reasoning teacher, but it is unclear which of those improvements survive when the teacher supervises a smaller on-policy student.…
cs.CL2026
Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator
Shiping Yang, Shining Liang, Weihao Liu +4
Identifying faithfulness hallucinations in LLM-generated outputs remains challenging due to the scarcity of high-quality annotated data. Recent work relies on advanced LLMs to synt…