Automated code completion, aiming at generating subsequent tokens from unfinished code, has significantly benefited from recent progress in pre-trained Large Language Models (LLMs). However, these models often suffer from coherence issues and hallucinations when dealing with complex code logic or extrapolating beyond their training data. Existing Retrieval Augmented Generation (RAG) techniques partially address these issues by retrieving relevant code with a separate encoding model where the retrieved snippet serves as contextual reference for code completion. However, their retrieval scope is subject to a singular perspective defined by the encoding model, which largely overlooks the complexity and diversity inherent in code semantics. To address this limitation, we propose ProCC, a code completion framework leveraging prompt engineering and the contextual multi-armed bandits algorithm to flexibly incorporate and adapt to multiple perspectives of code. ProCC first employs a prompt-based multi-retriever system which crafts prompt templates to elicit LLM knowledge to understand code semantics with multiple retrieval perspectives. Then, it adopts the adaptive retrieval selection algorithm to incorporate code similarity into the decision-making process to determine the most suitable retrieval perspective for the LLM to complete the code. Experimental results demonstrate that ProCC outperforms a widely studied code completion technique RepoCoder by 7.92% on the public benchmark CCEval, 3.19% in HumanEval-Infilling, 2.80% on our collected open-source benchmark suite, and 4.48% on the private-domain benchmark suite collected from Kuaishou Technology in terms of Exact Match. ProCC also allows augmenting fine-tuned techniques in a plug-and-play manner, yielding an averaged 6.5% improvement over the fine-tuned model.

Recent advances in LLM-based decompilers have been shown effective to convert low-level binaries into human-readable source code. However, there still lacks a comprehensive benchmark that provides large-scale binary-source function pairs, which is critical for advancing the LLM decompilation technology. Creating accurate binary-source mappings incurs severe issues caused by complex compilation settings and widespread function inlining that obscure the correspondence between binaries and their original source code. Previous efforts have either relied on used contest‐style benchmarks, synthetic binary–source mappings that diverge significantly from the mappings in real world, or partially matched binaries with only code lines or variable names, compromising the effectiveness of analyzing the binary functionality. To alleviate these issues, we introduce Decompile-Bench, the first open-source dataset comprising two million binary-source function pairs condensed from 100 million collected function pairs, i.e., 450GB of binaries compiled from permissively licensed GitHub projects. For the evaluation purposes, we also developed a benchmark Decompile-Bench-Eval including manually crafted binaries from the well-established HumanEval and MBPP, alongside the compiled GitHub repositories released after 2025 to mitigate data leakage issues. We further explore commonly-used evaluation metrics to provide a thorough assessment of the studied LLM decompilers and find that fine-tuning with Decompile-Bench causes a 20% improvement over previous benchmarks in terms of the re-executability rate. Our code and data has been released in HuggingFace and Github. https://github.com/anonepo/LLM4Decompile

LLMs demonstrate significant potential across various software engineering tasks. However, they still face challenges in generating correct code on the first attempt when addressing complex requirements. Introducing a discriminator to select reliable outputs from multiple generated results is an effective way to enhance their reliability and stability. Currently, these discriminators fall into two categories: execution-based discriminators and non-execution-based discriminators. Execution-based discriminators face flexibility challenges due to difficulties in obtaining test cases and security concerns, while non-execution-based discriminators, although more flexible, struggle to capture subtle differences in code details. To maintain flexibility while improving the model’s ability to capture fine-grained code details, this paper proposes Condor. We first design contrastive learning to optimize the code representations of the base model, enabling it to reflect differences in code details. Then, we leverage intermediate data from the code modification process to further enrich the discriminator’s training data, enhancing its ability to discern code details. Experimental results indicate that on the subtle code difference dataset (i.e., CodeNanoFix), Condor significantly outperforms other discriminators in discriminative performance: Condor (1.3B) improves the discriminative F1 score of DeepSeek-Coder (1.3B) from 67% to 73%. In discriminating LLM-generated outputs, Condor (1.3B) and Condor (110M) raise the Pass@1 score of Llama-3.1-Instruct (70B) on the CodeNanoFix dataset from 52.64% to 62.63% and 59.64%, respectively. Moreover, Condor demonstrates strong generalization capabilities on the APPS, MBPP, and LiveCodeBench datasets. For example, Condor (1.3B) improves the Pass@1 of Llama-3.1-Instruct (70B) on the APPS dataset by 147.05%.

Decompilation aims to convert binary code to high-level source code, but traditional tools like Ghidra often produce results that are difficult to read and execute.Motivated by the advancements in Large Language Models (LLMs), we propose LLM4Decompile, the first and largest open-source LLM series (1.3B to 33B) trained to decompile binary code.We optimize the LLM training process and introduce the LLM4Decompile-End models to decompile binary directly.The resulting models significantly outperform GPT-4o and Ghidra on the HumanEval and ExeBench benchmarks over 100% in terms of re-executability rate.Additionally, we improve the standard refinement approach to fine-tune the LLM4Decompile-Ref models, enabling them to effectively refine the decompiled code from Ghidra and achieve a further 16.2% improvement over the LLM4Decompile-End.LLM4Decompile 1 demonstrates the potential of LLMs to revolutionize binary code decompilation, delivering remarkable improvements in readability and executability while complementing conventional tools for optimal results.

For coverage-guided fuzzers, many of their adopted seeds are usually underused by exploring limited program states since essentially all their executions have to abide by rigorous program dependencies while only limited seeds are capable of accessing dependencies. Moreover, even when iteratively executing such limited seeds, the fuzzers have to repeatedly access the covered program states before uncovering new states. Such facts indicate that exploration power on program states of seeds has not been sufficiently leveraged by the existing coverage-guided fuzzing strategies. To tackle these issues, we propose a coverage-guided fuzzer, namely MirageFuzz, to mitigate the program dependencies when executing seeds for enhancing their exploration power on program states. Specifically, MirageFuzz first creates a “phantom” program of the target program by reducing its program dependencies corresponding to conditional statements while retaining their original semantics. Accordingly, MirageFuzz performs dual fuzzing, i.e., the source fuzzing to fuzz the original program and the phantom fuzzing to fuzz the phantom program simultaneously. Then, MirageFuzz applies the taint-based mutation mechanism to generate a new seed by updating the target conditional statement of a given seed from the source fuzzing with the corresponding condition value derived by the phantom fuzzing. To evaluate the effectiveness of MirageFuzz, we build a benchmark suite with 18 projects commonly adopted by recent fuzzing papers, and select seven open-source fuzzers as baselines for performance comparison with MirageFuzz. The experiment results suggest that MirageFuzz outperforms our baseline fuzzers from 13.42% to 77.96% averagely. Furthermore, MirageFuzz exposes 29 previously unknown bugs where 4 of them have been confirmed and 3 have been fixed by the corresponding developers.