2 papers
cs.CL2025
Language Models Can Learn from Verbal Feedback Without Scalar Rewards
Renjie Luo, Zichen Liu, Xiangyan Liu +5
LLMs are often trained with RL from human or AI feedback, yet such methods typically compress nuanced feedback into scalar rewards, discarding much of their richness and inducing s…
cs.CL2025
Through the Valley: Path to Effective Long CoT Training for Small Language Models
Renjie Luo, Jiaxi Li, Chen Huang +1
Long chain-of-thought (CoT) supervision has become a common strategy to enhance reasoning in language models. While effective for large models, we identify a phenomenon we call Lon…