Yuqi Tang

Vision-Language Introspection: Mitigating Overconfident Hallucinations in MLLMs via Interpretable Bi-Causal Steering

Shuliang Liu, Songbo Yang, Dong Fang, Sihang Jia et autres

Object hallucination critically undermines the reliability of Multimodal Large Language Models, often stemming from a fundamental failure in cognitive introspection, where models blindly trust linguistic priors over specific visual evidence. Existing mitigations remain limited: contrastive decoding approaches operate superficially without rectifying internal …

Vision-Language Introspection: Mitigating Overconfident Hallucinations in MLLMs via Interpretable Bi-Causal Steering

Liu S, Songbo Yang, Dong Fang, Sihang Jia et autres

Object hallucination critically undermines the reliability of Multimodal Large Language Models, often stemming from a fundamental failure in cognitive introspection, where models blindly trust linguistic priors over specific visual evidence. Existing mitigations remain limited: contrastive decoding approaches operate superficially without rectifying internal …

hk, cn (code pays fourni par la source)

Eliminating backdoor attacks on Vision Kolmogorov-Arnold networks via fine-function pruning

Yuqi Tang, Bo YUAN, Zhengpeng Chen, Jie Yang et autres

cn (code pays fourni par la source)

ClinDEF: A Dynamic Evaluation Framework for Large Language Models in Clinical Reasoning

Yuqi Tang, Jing Yu, Zichang Su, Kehua Feng et autres

Clinical diagnosis begins with doctor-patient interaction, during which physicians iteratively gather information, determine examination and refine differential diagnosis through patients' response. This dynamic clinical-reasoning process is poorly represented by existing LLM benchmarks that focus on static question-answering. To mitigate these gaps, recent …