LogitsCoder：通过logits偏好解码实现高效的思维链搜索以提升代码生成
LogitsCoder: Towards Efficient Chain-of-Thought Path Search via Logits Preference Decoding for Code Generation
AI总结：
LogitsCoder通过logit级控制机制提升代码生成的推理效率和质量，实现高效且有效的思维链搜索。
摘要：
代码生成仍是一个具有挑战性的任务，需要精确且结构化的推理。现有的测试时间扩展（TTS）方法，包括结构化树搜索，已在探索推理路径方面取得了进展，但仍面临两大主要挑战：（1）欠思，即推理链往往较浅，无法捕捉问题的全部复杂性；（2）过度思考，即过于冗长的推理导致效率低下和计算成本增加。为了解决这些问题，我们提出了LogitsCoder，一种新的框架，通过轻量级、logit级控制机制增强推理链。LogitsCoder通过首先通过logits偏好解码引导token选择向统计上更受欢迎的模式，然后使用logits排名路径选择和想法聚合选择和聚合多样化的推理路径。这导致了连贯且有效的推理链，平衡了深度和效率。广泛的实验表明，LogitsCoder生成的推理链更加高效且质量更高，相比基线方法在代码生成性能上表现更优。
英文摘要：
Code generation remains a challenging task that requires precise and structured reasoning. Existing Test Time Scaling (TTS) methods, including structured tree search, have made progress in exploring reasoning paths but still face two major challenges: (1) underthinking, where reasoning chains tend to be shallow and fail to capture the full complexity of problems; and (2) overthinking, where overly verbose reasoning leads to inefficiency and increased computational costs. To address these issues, we propose LogitsCoder, a novel framework that enhances chain-of-thought reasoning through lightweight, logit-level control mechanisms for code generation. LogitsCoder iteratively generates and refines reasoning steps by first steering token selection toward statistically preferred patterns via Logits Preference Decoding, then selecting and aggregating diverse reasoning paths using Logits Rank Based Path Selection and Thoughts Aggregation. This results in coherent and effective reasoning chains that balance depth and efficiency. Extensive experiments demonstrate that LogitsCoder produces more efficient and higher-quality reasoning chains, leading to superior code generation performance compared to baseline methods.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
