LAPO is a two-stage RL method that reduces reasoning token usage by up to 40.9% while slightly improving accuracy, by learning per-problem length targets from the model's own correct answers and prompting the model to declare its own budget.
Hapo: Training language models to reason concisely via history-aware policy optimization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
LAPO: Internalizing Reasoning Efficiency via Length-Adaptive Policy Optimization
LAPO is a two-stage RL method that reduces reasoning token usage by up to 40.9% while slightly improving accuracy, by learning per-problem length targets from the model's own correct answers and prompting the model to declare its own budget.