collaborators

21 papers

cs.LG2026

Towards Steering without Sacrifice: Principled Training of Steering Vectors for Prompt-only Interventions

Yuntai Bao, Qinfeng Li, Xinyan Yu +6

Recently, steering vectors (SVs) have emerged as an effective and lightweight approach to steer behaviors of large language models (LLMs), among which fine-tuned SVs are more effec…

cs.LG2026

Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding

Yingxuan Zhuang, Jingxiao Yang, Miao Pan +7

MLLMs frequently hallucinate objects inconsistent with visual inputs. This issue is typically attributed to the over-reliance on language priors, which can override the visual cont…

cs.AI2026

Domain-Specific Data Synthesis for LLMs via Minimal Sufficient Representation Learning

Tong Ye, Hang Yu, Tengfei Ma +6

Large Language Models have demonstrated remarkable progress in general-purpose capabilities and can achieve strong performance in specific domains through fine-tuning on domain-spe…

cs.CR2026

CoreGuard: Safeguarding Foundational Capabilities of LLMs Against Model Stealing in Edge Deployment

Qinfeng Li, Tianyue Luo, Xuhong Zhang +8

Proprietary large language models (LLMs) exhibit strong generalization capabilities across diverse tasks and are increasingly deployed on edge devices for efficiency and privacy re…

cs.CV2026

IBISAgent: Reinforcing Pixel-Level Visual Reasoning in MLLMs for Universal Biomedical Object Referring and Segmentation

Yankai Jiang, Qiaoru Li, Binlu Xu +6

Recent research on medical MLLMs has gradually shifted its focus from image-level understanding to fine-grained, pixel-level comprehension. Although segmentation serves as the foun…

cs.LG2026

Faithful Bi-Directional Model Steering via Distribution Matching and Distributed Interchange Interventions

Yuntai Bao, Xuhong Zhang, Jintao Chen +7

Intervention-based model steering offers a lightweight and interpretable alternative to prompting and fine-tuning. However, by adapting strong optimization objectives from fine-tun…