Shikun Zhang

TrustJudge: Inconsistencies of LLM-as-a-Judge and How to Alleviate Them

Autoformalizer with Tool Feedback

Decoupling Safety into Orthogonal Subspace: Cost-Efficient and Performance-Preserving Alignment for Large Language Models

Rethinking Deep Safety Alignment: Reflective Safety Alignment for Balancing Harmlessness and Helpfulness of LLMs

RewardAnything: Generalizable Principle-Following Reward Models

Decoupling Reasoning and Perception: An LLM-LMM Framework for Faithful Visual Reasoning

VLM-R³: Region Recognition, Reasoning, and Refinement for Enhanced Multimodal Chain-of-Thought

HaDeMiF: Hallucination Detection and Mitigation in Large Language Models

SAEMark: Steering Personalized Multilingual LLM Watermarks with Sparse Autoencoders

Boosting Resilience of Large Language Models through Causality-Driven Robust Optimization

Reasoning Through Execution: Unifying Process and Outcome Rewards for Code Generation

GETMusic: Generating Music Tracks with a Unified Representation and Diffusion Framework