AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

INTER: Mitigating Hallucination in Large Vision-Language Models by Interaction Guidance Sampling

Xin Dong, Shichao Dong, Jin Wang, Jing Huang, Li Zhou, Zenghui Sun, Lihua Jing, Jinsong Lan, Xiaoyong Zhu, Bo Zheng

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Preference tuning against hallucination · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply INTER on top of nucleus samplingInstructBLIP · MME totaltotal score: 1251.8 -> 1595.5Improved
apply INTER on top of nucleus samplingLLaVA-v1.5 7B · POPE COCOavg F1: 79.7 -> 85.7Improved
apply INTER on top of nucleus samplingLLaVA-v1.5 7B · CHAIR 64 tokensCHAIR_S: 24.4 -> 19.8Improved
apply INTER on top of OPERAInstructBLIP · CHAIR 64 tokensCHAIR_I: 6.8 -> 7.9Hurt
apply INTER on top of OPERAQwen-VL · POPE AOKVQAavg F1: 85.1 -> 84.4Hurt
apply INTER on top of VCDInstructBLIP · POPE AOKVQAavg F1: 81.0 -> 80.8Hurt
apply INTER on top of beam searchLLaVA-v1.5 7B · MM-Benchoverall accuracy: 65.1 -> 65.1No clear effect
apply INTER on top of nucleus samplingLLaVA-v1.5 7B · MM-Benchoverall accuracy: 57.3 -> 62.6Improved
apply INTER on top of greedy searchmPLUG-owl2 · MM-Benchoverall accuracy: 63.5 -> 62.3Hurt
add Interactive Guided Locator to Interaction Probability Modifier (ablation)INTER · MME total (LLaVA-v1.5 7B)total score: 1648.6 -> 1744.0Improved
apply modifier at all steps (k = 0) instead of only key tokens (k = 1) (ablation)INTER · CHAIR long captionsCHAIRHurt
vary threshold k within 0.5 to 1.5 (ablation)INTER · various LVLMsperformanceNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy