paper-with-me

Papers

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

2024-12-30 · Xiaoqing Zhang, YuHan Liu, Flood Sung, Xiuying Chen, Shuo Shang, Rui Yan

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 \textbf{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.

📄 PDF Abstract BibTeX arXiv:2502.17442

Code (0)

등록된 구현이 없습니다.

Tasks

Code GenerationHumanEvalmbpp

Methods 이 논문이 사용한 방법론

BASE 설명 없음

Similar Papers 제목 키워드 기반

An Empirical Study of Reasoning Steps in Thinking Code LLMs

2025-11-08 · Haoran Xue, Gias Uddin, Song Wang arxiv

Thinking Large Language Models (LLMs) generate explicit intermediate reasoning traces before final answers, potentially improving transparency, interpretability, and solution accuracy for code generation. However, the qu…

Code Generation

Think Anywhere in Code Generation

2026-03-31 · Xue Jiang, Tianyu Zhang, Ge Li, Mengyang Liu 외 arxiv

Recent advances in reasoning Large Language Models (LLMs) have primarily relied on upfront thinking, where reasoning occurs before final answer. However, this approach suffers from critical limitations in code generation…

Code Generation

Mellum2 Technical Report

2026-05-29 · Marko Kojic, Ivan Bondyrev, Aral de Moor, Joseph Shtok 외 arxiv

We present Mellum 2, an open-weight 12B-parameter Mixture-of-Experts (MoE) language model with 2.5B active parameters per token. Mellum 2 is a general-purpose language model specialized in software engineering, spanning …

Code Generation

Thinking vs. Doing: Agents that Reason by Scaling Test-Time Interaction

2025-06-09 · Junhong Shen, Hao Bai, Lunjun Zhang, Yifei Zhou 외

The current paradigm of test-time scaling relies on generating long reasoning traces ("thinking" more) before producing a response. In agent problems that require interaction, this can be done by generating thinking trac…

Reinforcement Learning (RL)

OctoNav: Towards Generalist Embodied Navigation

2025-06-11 · Chen Gao, Liankai Jin, Xingyu Peng, Jiazhao Zhang 외

Embodied navigation stands as a foundation pillar within the broader pursuit of embodied AI. However, previous navigation research is divided into different tasks/capabilities, e.g., ObjNav, ImgNav and VLN, where they di…