Guide, a new RL method that injects hints on total failure with off-policy correction, improves math pass@1 by a few points, while the paper argues RLVR progress is mostly self-distillation.
Openai’s reinforcement fine-tuning research program, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
REJECT 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Adaptive Guidance Accelerates Reinforcement Learning of Reasoning Models
Guide, a new RL method that injects hints on total failure with off-policy correction, improves math pass@1 by a few points, while the paper argues RLVR progress is mostly self-distillation.