Lune

ACL Findings2025

Thinking Before Running! Efficient Code Generation with Thorough Exploration and Optimal Refinement

Xiaoqing Zhang, Yuhan Liu, Flood Sung, Xiuying Chen, Shuo Shang, Rui Yan

Abstract

Code generation is crucial in software engineering for automating the coding process efficiently. While test-time computation methods show promise, they suffer from high latency due to multiple computation rounds. To overcome this, we introduce ThinkCoder, a framework that combines thorough exploration with optimal refinement. The exploration phase diversifies the solution space by searching for potential solutions, followed by a refinement phase that enhances precision. This approach allows us to select the best solution through careful consideration before taking action, avoiding excessive trial and error. To further minimize test-time computation overhead, we introduce preference-driven optimization with Reinforced Self-Training (ReST), which uses exploration trajectories from ThinkCoder to guide LLM's evolution. This approach enhances LLM's exploration efficiency via preference learning, cutting costs while maintaining accuracy. ThinkCoder boosts the performance with a single LLM, excelling on benchmarks like HumanEval and MBPP. Compared to SOTA models, it improves Pass@1 by 3.0% over MapCoder with just 6.4% of the computation cost. Against AgentCoder, ThinkCoder achieves a 0.5% higher Pass@1 after 2 rounds, outperforming AgentCoder's 5 rounds. Additionally, ReST with success trajectories enhances efficiency, allowing models like LLaMA2-7B to achieve competitive results using only 20% of the computational resources. These results highlight the framework's effectiveness and scalability. 1

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

1 top-tier paper has cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step by Step cites OctoPack: Instruction Tuning Code Large Language Models
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers1

  1. The Truth Becomes Clearer Through Debate! Multi-Agent Systems with Large Language Models Unmask Fake NewsYuhan Liu, Yuxuan Liu, Xiaoqing Zhang, Xiuying Chen et al.SIGIR 2025 · 20 citations

Builds on11

  1. LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu et al.ICLR 2022 · 18,833 citations
  2. ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIsYujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu et al.ICLR 2024 · 1,469 citations
  3. OctoPack: Instruction Tuning Code Large Language ModelsNiklas Muennighoff, Qian Liu, Armel Randy Zebaze, Qinkai Zheng et al.ICLR 2024 · 203 citations
  4. Exploring the Potential of ChatGPT in Automated Code Refinement: An Empirical StudyQi Guo, Junming Cao, Xiaofei Xie, Shangqing Liu et al.ICSE 2024 · 107 citations
  5. Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step by StepLi Zhong, Zilong Wang, Jingbo ShangACL Findings 2024 · 69 citations

Related papers

  1. ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement LearningJuyong Jiang, Jiasi Shen, Sunghun Kim, Kang Min Yoo et al.ACL Findings 2026 · 2 citations
  2. PairCoder: Pair Programming-Inspired Two-Agent Collaboration for Code GenerationJunhao Chen, Xiang Li, Yibin Xu, Yuehan Cui et al.ACL Findings 2026 · 2 citations
  3. EffiLearner: Enhancing Efficiency of Generated Code via Self-OptimizationDong Huang, Jianbo Dai, Han Weng, Puzhen Wu et al.NeurIPS 2024 · 54 citations
  4. A Pair Programming Framework for Code Generation via Multi-Plan Exploration and Feedback-Driven RefinementHuan Zhang, Wei Cheng, Yuhan Wu, Wei HuASE 2024 · 7 citations
  5. ACECODER: Acing Coder RL via Automated Test-Case SynthesisHuaye Zeng, Dongfu Jiang, Haozhe Wang, Ping Nie et al.ACL 2025 · 72 citations
Dusk over the sea between two cliffs drawn in fine vertical lines