Chong Wang

Junwei Liu, Chen Xu, Chong Wang et al.

Recent advances in large language model agents offer the promise of automating end-to-end software development from natural language requirements. However, existing approaches largely adopt linear, waterfall-style pipelines, which oversimplify the iterative nature of real-world development and struggle with complex, large-scale projects. To address these limitations, we propose EvoDev, an iterative software development framework inspired by feature-driven development. EvoDev decomposes user requirements into a set of user-valued features and constructs a Feature Map, a directed acyclic graph that explicitly models dependencies between features. Each node in the feature map maintains multi-level information, including business logic, design, and code, which is propagated along dependencies to provide context for subsequent development iterations. We evaluate EvoDev on challenging Android development tasks and show that it outperforms the best-performing baseline, Claude Code, by a substantial margin of 56.8%, while improving single-agent performance by 16.0%-76.6% across different base LLMs, highlighting the importance of dependency modeling, context propagation, and workflow-aware agent design for complex software projects. Our work summarizes practical insights for designing iterative, LLM-driven development frameworks and informs future training of base LLMs to better support iterative software development.

Linxi Liang, Jing Gong, Mingwei Liu et al.

Large Language Models (LLMs) have become pivotal tools for automating code generation in software development. However, these models face significant challenges in producing version-aware code for rapidly evolving languages like Rust, where frequent Application Programming Interfaces (API) changes across versions lead to compatibility issues and correctness errors. Existing benchmarks lack systematic evaluation of how models navigate API transitions, relying on labor-intensive manual curation and offering limited version-specific insights. To address this gap, we present RustEvo, a novel framework for constructing dynamic benchmarks that evaluate the ability of LLMs to adapt to evolving Rust APIs. RustEvo automates dataset creation by synthesizing 588 API changes (380 from Rust standard libraries, 208 from 15 third-party crates) into programming tasks mirroring real-world challenges. These tasks cover four API evolution categories: Stabilizations, Signature Changes, Behavioral Changes, and Deprecations, reflecting their actual distribution in the Rust ecosystem. Experiments on state-of-the-art (SOTA) LLMs reveal significant performance variations: models achieve a 65.8% average success rate on stabilized APIs but only 38.0% on behavioral changes, highlighting difficulties in detecting semantic shifts without signature alterations. Knowledge cutoff dates strongly influence performance, with models scoring 56.1% on before-cutoff APIs versus 32.5% on after-cutoff tasks. Retrieval-Augmented Generation (RAG) mitigates this gap, improving success rates by 13.5% on average for APIs released after model training. Our findings underscore the necessity of our evolution-aware benchmarks to advance the adaptability of LLMs in fast-paced software ecosystems. The framework and the benchmarks are publicly released at https://github.com/SYSUSELab/RustEvo.

Weisong Sun, Yuchen Chen, Mengzhe Yuan et al.

Neural code models (NCMs) have demonstrated extraordinary capabilities in code intelligence tasks. Meanwhile, the security of NCMs and NCMs-based systems has garnered increasing attention. In particular, NCMs are often trained on large-scale data from potentially untrustworthy sources, providing attackers with the opportunity to manipulate them by inserting crafted samples into the data. This type of attack is called a code poisoning attack (also known as a backdoor attack). It allows attackers to implant backdoors in NCMs and thus control model behavior, which poses a significant security threat. However, there is still a lack of effective techniques for detecting various complex code poisoning attacks. In this paper, we propose an innovative and lightweight technique for code poisoning detection named KillBadCode. KillBadCode is designed based on our insight that code poisoning disrupts the naturalness of code. Specifically, KillBadCode first builds a code language model (CodeLM) on a lightweight $n$-gram language model. Then, given poisoned data, KillBadCode utilizes CodeLM to identify those tokens in (poisoned) code snippets that will make the code snippets more natural after being deleted as trigger tokens. Considering that the removal of some normal tokens in a single sample might also enhance code naturalness, leading to a high false positive rate (FPR), we aggregate the cumulative improvement of each token across all samples. Finally, KillBadCode purifies the poisoned data by removing all poisoned samples containing the identified trigger tokens. The experimental results on two code poisoning attacks and four code intelligence tasks demonstrate that KillBadCode significantly outperforms four baselines. More importantly, KillBadCode is very efficient, with a minimum time consumption of only 5 minutes, and is 25 times faster than the best baseline on average.

Yanqing Wang, Chong Wang, Xiaojie Li et al.

With the rapid increasing of software project size and maintenance cost, adherence to coding standards especially by managing identifier naming, is attracting a pressing concern from both computer science educators and software managers. Software developers mainly use identifier names to represent the knowledge recorded in source code. However, the popularity and adoption consistency of identifier naming conventions have not been revealed yet in this field. Taking forty-eight popular open source projects written in three top-ranking programming languages Java, C and C++ as examples, an identifier extraction tool based on regular expression matching is developed. In the subsequent investigation, some interesting findings are obtained. For the identifier naming popularity, it is found that Camel and Pascal naming conventions are leading the road while Hungarian notation is vanishing. For the identifier naming consistency, we have found that the projects written in Java have a much better performance than those written in C and C++. Finally, academia and software industry are urged to adopt the most popular naming conventions consistently in their practices so as to lead the identifier naming to a standard, unified and high-quality road.

Renqin Cai, Qinglei Wang, Chong Wang et al.

Sequential recommendation recommends items based on sequences of users' historical actions. The key challenge in it is how to effectively model the influence from distant actions to the action to be predicted, i.e., recognizing the long-term dependence structure; and it remains an underexplored problem. To better model the long-term dependence structure, we propose a GatedLongRec solution in this work. To account for the long-term dependence, GatedLongRec extracts distant actions of top-$k$ related categories to the user's ongoing intent with a top-$k$ gating network, and utilizes a long-term encoder to encode the transition patterns among these identified actions. As user intent is not directly observable, we take advantage of available side-information about the actions, i.e., the category of their associated items, to infer the intents. End-to-end training is performed to estimate the intent representation and predict the next action for sequential recommendation. Extensive experiments on two large datasets show that the proposed solution can recognize the structure of long-term dependence, thus greatly improving the sequential recommendation.