9 papers
Long-Horizon Agent Trajectory Attribution: A Unified Benchmark and Fine-Grained Annotation Framework
Jing Chen, Yang Sun, Li Zhang +2
Large language model (LLM) agents increasingly operate through long-horizon trajectories involving user instructions, tool use, external observations, and memory. Existing benchmar…
Beyond Real Data: Synthetic Data through the Lens of Regularization
Amitis Shidani, Tyler Farghly, Yang Sun +2
Synthetic data can improve generalization when real data is scarce, but excessive reliance may introduce distributional mismatches that degrade performance. In this paper, we prese…
Discern Truth from Falsehood: Reducing Over-Refusal via Contrastive Refinement
Yuxiao Lu, Lin Xu, Yang Sun +2
Large language models (LLMs) aligned for safety often suffer from over-refusal, the tendency to reject seemingly toxic or benign prompts by misclassifying them as toxic. This behav…
LFD: Layer Fused Decoding to Exploit External Knowledge in Retrieval-Augmented Generation
Yang Sun, Zhiyong Xie, Lixin Zou +7
Retrieval-augmented generation (RAG) incorporates external knowledge into large language models (LLMs), improving their adaptability to downstream tasks and enabling information up…
AInsteinBench: Benchmarking Coding Agents on Scientific Repositories
Titouan Duston, Shuo Xin, Yang Sun +26
We introduce AInsteinBench, a large-scale benchmark for evaluating whether large language model (LLM) agents can operate as scientific computing development agents within real rese…
Causal-Guided Detoxify Backdoor Attack of Open-Weight LoRA Models
Linzhi Chen, Yang Sun, Hongru Wei +1
Low-Rank Adaptation (LoRA) has emerged as an efficient method for fine-tuning large language models (LLMs) and is widely adopted within the open-source community. However, the dece…