8 papers
One Model Is Enough: Native Retrieval Embeddings from LLM Agent Hidden States
Bo Jiang
LLM agents that retrieve external knowledge typically generate a search query as text, then run a separate embedding model to encode it into a vector. This two-model pipeline adds…
DistillGuard: Evaluating Defenses Against LLM Knowledge Distillation
Bo Jiang
Knowledge distillation from proprietary LLM APIs poses a growing threat to model providers, yet defenses against this attack remain fragmented and unevaluated. We present DistillGu…
KCoEvo: A Knowledge Graph Augmented Framework for Evolutionary Code Generation
Jiazhen Kang, Yuchen Lu, Chen Jiang +6
Code evolution is inevitable in modern software development. Changes to third-party APIs frequently break existing code and complicate maintenance, posing practical challenges for…
Benchmarking LLMs for Fine-Grained Code Review with Enriched Context in Practice
Ruida Hu, Xinchen Wang, Xin-Cheng Wen +5
Code review is a cornerstone of software quality assurance, and recent advances in Large Language Models (LLMs) have shown promise in its automation. However, existing benchmarks f…
CodeRepoQA: A Large-scale Benchmark for Software Engineering Question Answering
Ruida Hu, Chao Peng, Jingyi Ren +6
In this work, we introduce CodeRepoQA, a large-scale benchmark specifically designed for evaluating repository-level question-answering capabilities in the field of software engine…
DialogAgent: An Auto-engagement Agent for Code Question Answering Data Production
Xiaoyun Liang, Jingyi Ren, Jiayi Qi +2
Large Language Models (LLMs) have become increasingly integral to enhancing developer productivity, particularly in code generation, comprehension, and repair tasks. However, fine-…