19 papers
Semantic-Enhanced Automatic Refinement of Architecture Recovery Results Using LLMs
Yiran Zhang, Chengwei Liu, Yuqiang Sun +5
Understanding the architecture is crucial for effectively maintaining and managing large software systems. However, discrepancies often exist between the designed and implemented a…
DRIFT: Difficulty Routing Self-DIstillation with Rhythm-Gated Exploration and Success BuFfer Training
Haisen Luo, Yiwei Liu, Haoning Wang +13
Enabling large language models to achieve stable self-improvement without external expert supervision remains a central challenge in complex reasoning tasks. Existing self-distilla…
ATLAS: Agentic Taxonomy of Large-Scale Software Ecosystems
Junyi Lu, Mengyao Lyu, Jiahui Wu +6
The open-source ecosystem on GitHub lacks a systematic hierarchical taxonomy of software repositories. GitHub Topics, the dominant organizational mechanism, is flat, inconsistent,…
From Docs to Descriptions: Smell-Aware Evaluation of MCP Server Descriptions
Peiran Wang, Ying Li, Yuqiang Sun +3
The Model Context Protocol (MCP) has rapidly become a de facto standard for connecting LLM-based agents with external tools via reusable MCP servers. In practice, however, server s…
LogicScan: An LLM-driven Framework for Detecting Business Logic Vulnerabilities in Smart Contracts
Jiaqi Gao, Zijian Zhang, Yuqiang Sun +5
Business logic vulnerabilities have become one of the most damaging yet least understood classes of smart contract vulnerabilities. Unlike traditional bugs such as reentrancy or ar…
ASTRA: Automated Synthesis of agentic Trajectories and Reinforcement Arenas
Xiaoyu Tian, Haotian Wang, Shuaiting Chen +12
Large language models (LLMs) are increasingly used as tool-augmented agents for multi-step decision making, yet training robust tool-using agents remains challenging. Existing meth…