Saved in:
Bibliographic Details
Main Authors: Wang, Hao, Liu, Boyi, Zhang, Yufeng, Chen, Jie
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.12544
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913628578381824
author Wang, Hao
Liu, Boyi
Zhang, Yufeng
Chen, Jie
author_facet Wang, Hao
Liu, Boyi
Zhang, Yufeng
Chen, Jie
contents Competition-level code generation tasks pose significant challenges for current state-of-the-art large language models (LLMs). For example, on the LiveCodeBench-Hard dataset, models such as O1-Mini and O1-Preview achieve pass@1 rates of only 0.366 and 0.143, respectively. While tree search techniques have proven effective in domains like mathematics and general coding, their potential in competition-level code generation remains under-explored. In this work, we propose a novel token-level tree search method specifically designed for code generation. Leveraging Qwen2.5-Coder-32B-Instruct, our approach achieves a pass rate of 0.305 on LiveCodeBench-Hard, surpassing the pass@100 performance of GPT4o-0513 (0.245). Furthermore, by integrating Chain-of-Thought (CoT) prompting, we improve our method's performance to 0.351, approaching O1-Mini's pass@1 rate. To ensure reproducibility, we report the average number of generations required per problem by our tree search method on the test set. Our findings underscore the potential of tree search to significantly enhance performance on competition-level code generation tasks. This opens up new possibilities for large-scale synthesis of challenging code problems supervised fine-tuning (SFT) data, advancing competition-level code generation tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2412_12544
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Seed-CTS: Unleashing the Power of Tree Search for Superior Performance in Competitive Coding Tasks
Wang, Hao
Liu, Boyi
Zhang, Yufeng
Chen, Jie
Artificial Intelligence
Software Engineering
Competition-level code generation tasks pose significant challenges for current state-of-the-art large language models (LLMs). For example, on the LiveCodeBench-Hard dataset, models such as O1-Mini and O1-Preview achieve pass@1 rates of only 0.366 and 0.143, respectively. While tree search techniques have proven effective in domains like mathematics and general coding, their potential in competition-level code generation remains under-explored. In this work, we propose a novel token-level tree search method specifically designed for code generation. Leveraging Qwen2.5-Coder-32B-Instruct, our approach achieves a pass rate of 0.305 on LiveCodeBench-Hard, surpassing the pass@100 performance of GPT4o-0513 (0.245). Furthermore, by integrating Chain-of-Thought (CoT) prompting, we improve our method's performance to 0.351, approaching O1-Mini's pass@1 rate. To ensure reproducibility, we report the average number of generations required per problem by our tree search method on the test set. Our findings underscore the potential of tree search to significantly enhance performance on competition-level code generation tasks. This opens up new possibilities for large-scale synthesis of challenging code problems supervised fine-tuning (SFT) data, advancing competition-level code generation tasks.
title Seed-CTS: Unleashing the Power of Tree Search for Superior Performance in Competitive Coding Tasks
topic Artificial Intelligence
Software Engineering
url https://arxiv.org/abs/2412.12544