1 citations · 1 across the 16 of their papers we have counts for
7 papers · 1 filter
Exact Risk Certificates for Pairwise-Judged Reranking
Xinyu Wang, Jingrui Tian, Sicheng Lyu +4
Pairwise relevance judgments constrain both the loss numerator and the ideal-ranking denominator of nDCG. We characterize the exact worst-case loss of replacing a ranking with a fi…
Measuring the Task Value of Precision Upgrades in Closed-Loop Agents
Sicheng Lyu, Xinyu Wang, Yu Gu +3
A low-precision agent can call its full-precision (BF16) model at selected decisions as a fallback. A learned precision router chooses where to call it from the disagreement betwee…
From Mathematical to Executable Certificates for Machine Unlearning
Ziyu Zhao, Xinyu Wang, Xiaowen Chang +1
Machine unlearning is needed when data must be removed because of deletion requests, outdated records, or data-quality concerns, while retraining from scratch can be costly. Certif…
RRC: Unlocking Generative Reward Models in LLM Reinforcement Learning via Ranking-Based Reward Construction
Chenglong Wang, Ziming Zhu, Yifu Huo +9
Recent advances in reward modeling show a paradigm shift from discriminative reward models to generative reward models. However, despite their strong capabilities in response ranki…
OJBKQ: Objective-Joint Babai-Klein Quantization
Xinyu Wang, Ziyu Zhao, Peng Lu +2
Post-training quantization (PTQ) is widely used to compress large language models without retraining. However, many existing weight-only methods rely on heuristic objectives and gr…
Beyond Hard Writes and Rigid Preservation: Soft Recursive Least-Squares for Lifelong LLM Editing
Xinyu Wang, Sicheng Lyu, Yu Gu +4
Model editing updates a pre-trained LLM with new facts or rules without retraining while preserving unrelated behavior. In real deployment, edits arrive as long streams, creating a…
MARS: Unleashing the Power of Speculative Decoding via Margin-Aware Verification
Jingwei Song, Xinyu Wang, Hanbin Wang +6
Speculative Decoding (SD) accelerates autoregressive large language model (LLM) inference by decoupling generation and verification. While recent methods improve draft quality by t…