共计 146 篇文章

---

2026

07-04

Cross-Modal Attention Calibration for LVLM Hallucination Mitigation

07-04

First Logit Boosting： Visual Grounding Method to Mitigate Object Hallucination in Large Vision-Language Models

07-03

Train Once, Reuse Everywhere：Generalizable Implicit In-Context Learning by Routing Attention

07-02

ViT³： Unlocking Test-Time Training in Vision

07-01

ShellGames： Speculative LLM-Driven SSH Deception

06-23

Automatic Layer Selection for Hallucination Detection

06-22

Zone of Proximal Policy Optimization： Teacher in Prompts, Not Gradients

06-21

VLM³： Vision Language Models Are Native 3D Learners

06-20

When is Your LLM Steerable？

06-17

Structural Graph Probing of Vision-Language Models