Large Language Models (LLMs) have shown impressive abilities in code generation. Chain-of-Thought (CoT) prompting is the state-of-the-art approach to utilizing LLMs. CoT prompting asks LLMs first to generate CoTs (i.e., intermediate natural language reasoning steps) and then output the code. However, the accuracy of CoT prompting still cannot satisfy practical applications. For example, gpt-3.5-turbo with CoT prompting only achieves 53.29% Pass@1 in HumanEval. In this article, we propose Structured CoTs (SCoTs) and present a novel prompting technique for code generation named SCoT prompting. Our motivation is that human developers follow structured programming. Developers use three programming structures (i.e., sequential, branch, and loop) to design and implement structured programs. Thus, we ask LLMs to use three programming structures to generate SCoTs (structured reasoning steps) before outputting the final code. Compared to CoT prompting, SCoT prompting explicitly introduces programming structures and unlocks the structured programming thinking of LLMs. We apply SCoT prompting to two LLMs (i.e., gpt-4-turbo, gpt-3.5-turbo, and DeepSeek Coder-Instruct- \(\{\) 1.3B, 6.7B, 33B \(\}\) ) and evaluate it on three benchmarks (i.e., HumanEval, MBPP, and MBCPP). SCoT prompting outperforms CoT prompting by up to 13.79% in Pass@1. SCoT prompting is robust to examples and achieves substantial improvements. The human evaluation also shows human developers prefer programs from SCoT prompting.

Code translation tools, namely transpilers, are developed for automatic source-to-source translation. Latest learning-based transpilers have shown impressive enhancement against rule-based counterparts in both translation accuracy and readability, owing to their task-specific pre-training on extensive monolingual corpora. Nevertheless, their current performance still remains unsatisfactory for practical deployment, and the associated training resources are also prohibitively expensive. Large Language Models (LLMs), pre-trained on huge amounts of human-written code/text, have shown remarkable performance in many code intelligence tasks due to their powerful generality, even without task-specific re-training/fine-tuning. Thus, LLMs can potentially circumvent the above limitations, but they have not been exhaustively explored yet. This paper investigates diverse LLMs and learning-based transpilers for automated code translation tasks, finding that: although certain LLMs have outperformed current transpilers, they still have some accuracy issues, where most of the failures are induced by a lack of comprehension of source programs (38.51%), missing clear instructions on I/O types in translation (14.94%), and ignoring discrepancies between source and target programs (41.38%). Enlightened by the above findings, we further propose UniTrans , a Unified code Translation framework, applicable to various LLMs, for unleashing their power in this field. Specifically, UniTrans first crafts a series of test cases for target programs with the assistance of source programs. Next, it harnesses the above auto-generated test cases to augment the code translation and then evaluate their correctness via execution. Afterward, UniTrans further (iteratively) repairs incorrectly translated programs prompted by test case execution results. Extensive experiments are conducted on six settings of translation datasets between Python, Java, and C++. Three recent LLMs of diverse sizes, including GPT-3.5 and LLaMA-13B/7B, are tested with UniTrans , and all achieve substantial improvements in terms of computational accuracy and exact match accuracy among almost all translation settings, showing the universal effectiveness of UniTrans in practice.

Jia Li, Ge Li, Yunfei Zhao, Yongmin Li, Huanyu Liu, Hao Zhu, Lecheng Wang, Kaibo Liu, Zheng Fang, Lanshen Wang, Jiazheng Ding, Xuanming Zhang, Yuqi Zhu, Yihong Dong, Zhi Jin, Binhua Li, Fei Huang, Yongbin Li, Bin Gu, Mengfei Yang. Findings of the Association for Computational Linguistics ACL 2024. 2024.

Large Language Models (LLMs) have shown impressive In-Context Learning (ICL) ability in code generation. LLMs take a prompt context consisting of a few demonstration examples and a new requirement as input, and output new programs without any parameter update. Existing studies have found that the performance of ICL-based code generation heavily depends on the quality of demonstration examples and thus arises research on selecting demonstration examples: given a new requirement, a few demonstration examples are selected from a candidate pool, where LLMs are expected to learn the pattern hidden in these selected demonstration examples. Existing approaches are mostly based on heuristics or randomly selecting examples. However, the distribution of randomly selected examples usually varies greatly, making the performance of LLMs less robust. The heuristics retrieve examples by only considering textual similarities of requirements, leading to sub-optimal performance. To fill this gap, we propose a L arge language model- A ware selection approach for I n-context- L earning-based code generation named LAIL. LAIL uses LLMs themselves to select examples. It requires LLMs themselves to label a candidate example as a positive example or a negative example for a requirement. Positive examples are helpful for LLMs to generate correct programs, while negative examples are trivial and should be ignored. Based on the labeled positive and negative data, LAIL trains a model-aware retriever to learn the preference of LLMs and select demonstration examples that LLMs need. During the inference, given a new requirement, LAIL uses the trained retriever to select a few examples and feed them into LLMs to generate desired programs. We apply LAIL to four widely used LLMs and evaluate it on five code generation datasets. Extensive experiments demonstrate that LAIL outperforms the state-of-the-art (SOTA) baselines by 11.58%, 3.33%, and 5.07% on CodeGen-Multi-16B, 1.32%, 2.29%, and 1.20% on CodeLlama-34B, and achieves 4.38%, 2.85%, and 2.74% improvements on Text-davinci-003 in terms of Pass@1 at MBJP, MBPP, and MBCPP, respectively. In addition to function-level code generation, LAIL improves the performance of LLMs on DevEval, a repository-level code generation dataset, which achieves 10.04%, 8.12%, and 4.63% improvements compared to the SOTA baselines at Pass@1, 3, and 5 on CodeLlama-7B. Human evaluation further verifies that the generated programs of LAIL are superior in correctness, code quality, and maintainability. Besides, LAIL has satisfactory transferability across different LLMs and datasets, where the retriever learned on one LLM (dataset) can be transferred to other LLMs (datasets).

Developers often perform repetitive code editing activities (up to 70%) for various reasons (e.g., code refactoring) during software development. Many deep learning (DL) models have been proposed to automate code editing by learning from the code editing history. Among DL-based models, pre-trained code editing models have achieved the state-of-the-art (SOTA) results. Pre-trained models are first pre-trained with pre-training tasks and fine-tuned with the code editing task. Existing pre-training tasks mainly are code infilling tasks (e.g., masked language modeling), which are derived from the natural language processing field and are not designed for automatic code editing. In this paper, we propose a novel pre-training task specialized in code editing and present an effective pre-trained code editing model named CodeEditor. Compared to previous code infilling tasks, our pre-training task further improves the performance and generalization ability of code editing models. Specifically, we collect lots of real-world code snippets as the ground truth and use a powerful generator to rewrite them into mutated versions. Then, we pre-train our CodeEditor to edit mutated versions into the corresponding ground truth, to learn edit patterns. We conduct experiments on four code editing datasets and evaluate the pre-trained CodeEditor in three settings (i.e., fine-tuning, few-shot, and zero-shot). (1) In the fine-tuning setting, we train the pre-trained CodeEditor with four datasets and evaluate it on the test data. CodeEditor outperforms the SOTA baselines by 15%, 25.5%, and 9.4% and 26.6% on four datasets. (2) In the few-shot setting, we train the pre-trained CodeEditor with limited data and evaluate it on the test data. CodeEditor substantially performs better than all baselines, even outperforming baselines that are fine-tuned with all data. (3) In the zero-shot setting, we evaluate the pre-trained CodeEditor on the test data without training. CodeEditor correctly edits 1,113 programs while the SOTA baselines can not work. The results show that the superiority of our pre-training task and the pre-trained CodeEditor is more effective in automatic code editing.

Large language models (LLMs) have shown great success in code generation. LLMs take as the input a prompt and output the code. How to make prompts (i.e., Prompting Techniques ) is a key question. Existing prompting techniques are designed for natural language generation and have low accuracy in code generation. In this article, we propose a new prompting technique named AceCoder . Our motivation is that code generation meets two unique challenges (i.e., requirement understanding and code implementation). AceCoder contains two novel mechanisms (i.e., guided code generation and example retrieval) to solve these challenges. ❶ Guided code generation asks LLMs first to analyze requirements and output an intermediate preliminary (e.g., test cases). The preliminary clarifies requirements and tells LLMs “what to write.” ❷ Example retrieval selects similar programs as examples in prompts, which provide lots of relevant content (e.g., algorithms, APIs) and teach LLMs “how to write.” We apply AceCoder to four LLMs (e.g., GPT-3.5, CodeGeeX) and evaluate it on three public benchmarks using the Pass@ \(k\) . Results show that AceCoder can significantly improve the performance of LLMs on code generation. In terms of Pass@1, AceCoder outperforms the SOTA baseline by up to 56.4% in MBPP, 70.7% in MBJP, and 88.4% in MBJSP . AceCoder is effective in LLMs with different sizes (i.e., 6B–13B) and different languages (i.e., Python, Java, and JavaScript). Human evaluation shows human developers prefer programs from AceCoder .