الباحثون

Zhiliang Tian

المنشورات 2

نسخة أولية وصول مفتوح

Role-aware Heuristic Episodic Attention for Conversational LLMs

Wanyang Hong, Zhaoning Zhang, Yi Chen وآخرون · 2026

Large language models often lose track of persistent instructions and relevant information as multi-turn conversations grow. We study this cumulative contextual decay through three related failure modes: attention pollution, dilution, and drift. We propose REA (Role-aware Heuristic Episodic Attention), a context-manage …

نسخة أولية وصول مفتوح

Only Pay What You Must Spend: On-Demand Privacy Budget Payment for Differentially Private RAG

Zhonghao Sun, Zhiliang Tian, Xinyue Fang وآخرون · 2026

Deploying large language models (LLMs) on sensitive data via Retrieval-Augmented Generation (RAG) introduces severe privacy risks. Recent studies apply Differential Privacy (DP) to LLMs with RAG for formal privacy guarantees. However, existing DP-RAG frameworks rapidly exhaust the privacy budget. Although recent effort …

المؤلفون المشاركون