AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

See Only When Needed: Context-Aware Attention Intervention for Hallucination-Free LVLMs

Yuqing Lei, Wenbo Lyu, Yingjun Du, Xiantong Zhen, Cees Snoek, Ling Shao

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Decoding-time hallucination mitigation

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 CAI (token- and entropy-gated attention tilt) to LLaVA-1.5LLaVA-1.5 · CHAIR max 128CHAIR_S: 56.6 -> 39.6Improved
Attention-only CAI (lambda=1) without contrastive decoding (ablation)LLaVA-1.5 · POPE/CHAIR/MMEPOPE avg: 74.81 -> 77.13Improved
Uniform attention amplification (PAI†) instead of selective (ablation)LLaVA-1.5 · POPE/CHAIR/MMEPOPE avg: 74.81 -> 74.57No clear effect
Add contrastive decoding on top of CAI† (ablation)CAI · POPE avgaccuracy: 77.13 -> 83.67Improved
Apply CAI to InstructBLIPInstructBLIP · CHAIR max 128CHAIR_S: 54.2 -> 52.8No clear effect
Apply CAI on MMELLaVA-1.5 · MMEscore: 588.33 -> 660.00Improved
Apply CAI to Qwen3.5-4B softmax layers rather than GLA gates (ablation)Qwen3.5-4B · CHAIR max 128CHAIR_S: 35.0 -> 29.8Improved
Tune start layer ls and entropy threshold gamma (ablation)CAI · POPE A-OKVQA randomaccuracyMixed
Tune contrastive coefficient lambda (ablation)CAI · POPE A-OKVQA randomaccuracyMixed

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