Junjie Chen
5 papers · Latest:
Characterizing the Failure Modes of LLMs in Resolving Real-World GitHub Issues
This paper analyzes LLM failures in resolving GitHub issues, revealing strategy formulation as the most error-prone stage and localization as the least.
Characterizing and Mitigating False-Positive Bug Reports in the Linux Kernel
This paper characterizes false-positive bug reports in the Linux kernel and proposes LLM-based mitigation, showing they waste significant developer effort.
Improving LLM Code Generation via Requirement-Aware Curriculum Reinforcement Learning
RECRL improves LLM code generation by using a requirement-aware curriculum reinforcement learning framework for better training efficiency.
TEMPLATEFUZZ: Fine-Grained Chat Template Fuzzing for Jailbreaking and Red Teaming LLMs
TEMPLATEFUZZ is a novel fuzzing framework that systematically exploits vulnerabilities in LLM chat templates, achieving high jailbreak success rates with minimal accuracy loss.
REAgent: Requirement-Driven LLM Agents for Software Issue Resolution
REAgent is a requirement-driven LLM agent framework that improves software issue resolution by refining issue descriptions into structured requirements.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.