5 papers
Learning When Not to Act: Mitigating Tool Abuse in Agentic Reinforcement Learning
Liuji Chen, Dianxing Tang, Xing Shi +4
Agentic reinforcement learning can induce tool abuse, where models overuse external tools even for queries solvable by internal reasoning. Existing approaches mitigate this issue w…
From Profiles to Steering Vectors: Global Sparse Priors and Local Semantic Calibration for Personalized Text Generation
Liuji Chen, Zeyu Zhang, Xinyuan Zhang +4
Personalized text generation requires models to capture user-specific writing styles from historical data. Existing approaches based on retrieval, parameter-efficient fine-tuning,…
Graffe: Graph Representation Learning via Diffusion Probabilistic Models
Dingshuo Chen, Shuchen Xue, Liuji Chen +5
Diffusion probabilistic models (DPMs), widely recognized for their potential to generate high-quality samples, tend to go unnoticed in representation learning. While recent progres…
Uncovering Competing Poisoning Attacks in Retrieval-Augmented Generation
Liuji Chen, Xiaofang Yang, Yuanzhuo Lu +6
Retrieval-Augmented Generation (RAG) systems improve the factual grounding of large language models (LLMs) but remain vulnerable to retrieval poisoning, where adversaries seed the…
SEEM: Exploiting Black-Box Text Attacks to Manipulate Tool Selection
Liuji Chen, Hao Gao, Jinghao Zhang +3
Tool learning has emerged as a powerful auxiliary mechanism that extends the capabilities of large language models (LLMs), enabling them to address complex tasks that demand real-t…