共计 286 篇文章

---

2025

11-05

AGLA：Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attention

11-04

ICT：Image-Object Cross-Level Trusted Intervention for Mitigating Object Hallucination in Large Vision-Language Models

11-03

幻觉与越狱的一致性

11-02

少即是多：从 EOS 决策角度缓解多模态幻觉

11-01

性能提升的幻觉：为什么对比解码无法缓解多模态大模型中的对象幻觉问题？

10-31

基于参数高效微调的可扩展确切机器遗忘方法

10-29

频域特征对多模态大模型的幻觉影响

10-26

ChunkLLM：A Lightweight Pluggable Framework for Accelerating LLMs Inference

10-22

RoseLoRA：面向知识编辑与微调的预训练语言模型行与列稀疏低秩适配

10-21

拒绝如坠悬崖：推理中的安全对齐为何失效？