Papers › Efficient RL Training for Reasoning Models via Length-Aware Optimization

Efficient RL Training for Reasoning Models via Length-Aware Optimization

18 May 2025arXiv:2505.12284archive 2025-07-28

Danlong Yuan, Tian Xie, Shaohan Huang, Zhuocheng Gong, Huishuai Zhang, Chong Luo, Furu Wei, Dongyan Zhao

Large reasoning models, such as OpenAI o1 or DeepSeek R1, have demonstrated remarkable performance on reasoning tasks but often incur a long reasoning path with significant memory and time costs. Existing methods primarily aim to shorten reasoning paths by introducing additional training data and stages. In this paper, we propose three critical reward designs integrated directly into the reinforcement learning process of large reasoning models, which reduce the response length without extra training stages. Experiments on four settings show that our method significantly decreases response length while maintaining or even improving performance. Specifically, in a logic reasoning setting, we achieve a 40% reduction in response length averaged by steps alongside a 14% gain in performance. For math problems, we reduce response length averaged by steps by 33% while preserving performance.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

lblankl/short-rl officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Math

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections