paper-with-me

Papers

Afterburner: Reinforcement Learning Facilitates Self-Improving Code Efficiency Optimization

2025-05-29 · Mingzhe Du, Luu Anh Tuan, Yue Liu, Yuhao QING, Dong Huang, Xinyi He, Qian Liu, Zejun Ma, See-Kiong Ng

Large Language Models (LLMs) generate functionally correct solutions but often fall short in code efficiency, a critical bottleneck for real-world deployment. In this paper, we introduce a novel test-time iterative optimization framework to address this, employing a closed-loop system where LLMs iteratively refine code based on empirical performance feedback from an execution sandbox. We explore three training strategies: Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Group Relative Policy Optimization (GRPO). Experiments on our Venus dataset and the APPS benchmark show that SFT and DPO rapidly saturate in efficiency gains. In contrast, GRPO, using reinforcement learning (RL) with execution feedback, continuously optimizes code performance, significantly boosting both pass@1 (from 47% to 62%) and the likelihood of outperforming human submissions in efficiency (from 31% to 45%). Our work demonstrates effective test-time code efficiency improvement and critically reveals the power of RL in teaching LLMs to truly self-improve code efficiency.

📄 PDF Abstract BibTeX arXiv:2505.23387

Code (0)

등록된 구현이 없습니다.

Tasks

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

Methods 이 논문이 사용한 방법론

SFT Shrink and Fine-Tune, or SFT, is a type of distillation that avoids explicit distillation by copying parameters to a student student model and then fine-tuning.…
DPO 설명 없음

Similar Papers 제목 키워드 기반

MOOSS: Mask-Enhanced Temporal Contrastive Learning for Smooth State Evolution in Visual Reinforcement Learning

2024-09-02 · Jiarui Sun, M. Ugur Akcal, Wei zhang, Girish Chowdhary

In visual Reinforcement Learning (RL), learning from pixel-based observations poses significant challenges on sample efficiency, primarily due to the complexity of extracting informative state representations from high-d…

Contrastive Learninggraph constructionReinforcement Learning (RL)

Mask-based Latent Reconstruction for Reinforcement Learning

2022-01-28 · Tao Yu, Zhizheng Zhang, Cuiling Lan, Yan Lu 외

For deep reinforcement learning (RL) from pixels, learning effective state representations is crucial for achieving high performance. However, in practice, limited experience and high-dimensional inputs prevent effective…

Deep Reinforcement Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)+1

Multi-Task Reinforcement Learning for Quadrotors

2024-12-17 · Jiaxu Xing, Ismail Geles, Yunlong Song, Elie Aljalbout 외

Reinforcement learning (RL) has shown great effectiveness in quadrotor control, enabling specialized policies to develop even human-champion-level performance in single-task scenarios. However, these specialized policies…

Autonomous Racingreinforcement-learningReinforcement LearningReinforcement Learning (RL)+1

Learning State Representations via Retracing in Reinforcement Learning

2021-11-24 · ICLR 2022 4 · Changmin Yu, Dong Li, Jianye Hao, Jun Wang 외

We propose learning via retracing, a novel self-supervised approach for learning the state representation (and the associated dynamics model) for reinforcement learning tasks. In addition to the predictive (reconstructio…

continuous-controlContinuous ControlModel-based Reinforcement Learningreinforcement-learning+3

Efficient Training in Multi-Agent Reinforcement Learning: A Communication-Free Framework for the Box-Pushing Problem

2024-11-19 · David Ge, Hao Ji

Self-organizing systems consist of autonomous agents that can perform complex tasks and adapt to dynamic environments without a central controller. Prior research often relies on reinforcement learning to enable agents t…

Multi-agent Reinforcement Learning