AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

HIVE: Understanding Post Hallucination Reasoning in Vision Language Models

Feng He, Zhenting Wang, Qifan Wang, Qiang Guan, Dongfang Liu, Ruixiang Tang, Qiankun Li

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Decoding-time hallucination mitigation · 0 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
replace faithful caption with hallucinated captionGPT-4o · PlantVillageaccuracy: 62.73 -> 77.41Improved
replace faithful caption with hallucinated captionQwen-VL-Max · ISICaccuracy: 58.71 -> 75.62Improved
replace faithful caption with hallucinated captionQwen-VL-Max · GQAaccuracy: 69.88 -> 67.90Hurt
replace faithful caption with hallucinated captionGPT-4o, text-only tasks · BBBPaccuracy: 61.67 -> 68.33Improved
replace faithful caption with hallucinated captionO3 and DeepSeek R1, text-only tasks · BBBPaccuracy: 73.27 -> 59.47Hurt
replace faithful caption with hallucinated captionGPT-4o · CodeXGLUEaccuracy: 55.15 -> 52.75No clear effect
replace hallucination discriminator with random checker (ablation)HIVE · ISICaccuracy: 0.5763 -> 0.5805No clear effect
mask hallucinated evidence tokens in helpful captions (ablation)HIVE · AntiCP2accuracy: 0.244Improved
set sampling temperature to 0.6 instead of 0.0 (ablation)hallucinated caption generation · PlantVillageaccuracy gain (H-F): +2.30 -> +14.68Improved
raise temperature to 0.9 instead of 0.6 (ablation)hallucinated caption generation · AntiCP2accuracy gain (H-F): +3.76 -> -5.00Hurt
use 256 token budget instead of 128 (ablation)hallucinated caption generation · PlantVillageaccuracy gain (H-F): -4.66 -> +14.68Improved
use 1024 token budget instead of 256 (ablation)hallucinated caption generation · ISICaccuracy gain (H-F): +11.76 -> -0.48Hurt

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