EMNLP Findings2024
Rethinking Code Refinement: Learning to Judge Code Efficiency
摘要
Large Language Models (LLMs) have demonstrated impressive capabilities in understanding and generating codes. Due to these capabilities, many recent methods are proposed to automatically refine the codes with LLMs. However, we should rethink that the refined codes (from LLMs and even humans) are not always more efficient than their original versions. On the other hand, running two different versions of codes and comparing them every time is not ideal and time-consuming. Therefore, in this work, we propose a novel method based on the code language model that is trained to judge the efficiency between two different codes (generated across humans and machines) by either classifying the superior one or predicting the relative improvement. We validate our method on multiple programming languages with multiple refinement steps, demonstrating that the proposed method can effectively distinguish between more and less efficient versions of code.
问 Lune
几分钟读透这篇 Paper,不用再花一下午。
Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。
脉络
- CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis 引用了 Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Learning Performance-Improving Code Edits 引用了 CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis
- Learning Performance-Improving Code Edits 引用了 Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
它引用的 Paper5
- Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma 等NeurIPS 2022 · 22,562 次引用
- Unsupervised Translation of Programming LanguagesBaptiste Rozière, Marie-Anne Lachaux, Lowik Chanussot, Guillaume LampleNeurIPS 2020 · 606 次引用
- CodeGen: An Open Large Language Model for Code with Multi-Turn Program SynthesisErik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu 等ICLR 2023 · 234 次引用
- Learning Performance-Improving Code EditsAlexander Shypula, Aman Madaan, Yimeng Zeng, Uri Alon 等ICLR 2024 · 141 次引用
- CodeBERTScore: Evaluating Code Generation with Pretrained Models of CodeShuyan Zhou, Uri Alon, Sumit Agarwal, Graham NeubigEMNLP 2023 · 77 次引用
相关 Paper
- More Than Just Functional: LLM-as-a-Critique for Efficient Code GenerationDerui Zhu, Dingfan Chen, Jinfu Chen, Jens Grossklags 等NeurIPS 2025 · 2 次引用
- Generating Energy-Efficient Code via Large-Language Models - Where are we now?Radu Apsan, Vincenzo Stoico, Michel Albonico, Rudra Dhar 等ICSE 2026
- ECCO: Can We Improve Model-Generated Code Efficiency Without Sacrificing Functional Correctness?Siddhant Waghjale, Vishruth Veerendranath, Zhiruo Wang, Daniel FriedEMNLP 2024 · 3 次引用
- CYCLE: Learning to Self-Refine the Code GenerationYangruibo Ding, Marcus J. Min, Gail E. Kaiser, Baishakhi RayOOPSLA 2024 · 35 次引用
- Self-Edit: Fault-Aware Code Editor for Code GenerationKechi Zhang, Zhuo Li, Jia Li, Ge Li 等ACL 2023 · 42 次引用
