AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Entropy-Gradient Grounding: Training-Free Evidence Retrieval in Vision-Language Models

Marcel Gröpl, Jaewoo Jung, Seungryong Kim, Marc Pollefeys, Sunghwan Hong

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 4 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
use spatial-entropy stopping instead of confidence stopping (ablation)Ours on LLaVA-1.6 · V*accuracy: 70.16 -> 73.3Improved
use spatial-entropy stopping instead of 1 fixed iteration (ablation)Ours on LLaVA-1.6 · V*accuracy: 60.73 -> 73.3Improved
use spatial-entropy stopping instead of 3 fixed iterations (ablation)Ours on LLaVA-1.6 · DocVQAaccuracy: 61.80 -> 65.07Improved
add iterative refinement with stopping (4.80 s) versus single pass (ablation)Ours on LLaVA-1.6 · RWQAinference time: 2.22 s -> 4.80 sHurt
backpropagate Shannon entropy instead of max-probability log (ablation)Ours on LLaVA-1.6 · DocVQAaccuracy: 64.53 -> 65.07Improved
use top-P entropy instead of full entropy (ablation)Ours on LLaVA-1.6 · V*accuracy: 73.30 -> 72.25No clear effect
select 2 regions instead of 1 (ablation)Ours on LLaVA-1.6 · DocVQAaccuracy: 61.25 -> 65.07Improved
select 2 regions instead of 1 (ablation)Ours on LLaVA-1.6 · V*accuracy: 69.63 -> 73.30Improved
increase regions from 2 to 4 (ablation)Ours on LLaVA-1.6 · V*accuracy: 73.30 -> 72.25No clear effect
compute gradient at last layer 32 instead of layer 15 (ablation)Ours on LLaVA-1.6 · DocVQAaccuracy: 50.99 -> 65.07Improved
compute gradient at layer 20 instead of last layer (ablation)Ours on LLaVA-1.6 · V*accuracy: 73.30 -> 78.01Mixed
backpropagate from first token instead of second (ablation)Ours on LLaVA-1.6 · TextVQAaccuracy: 65.27 -> 67.96Improved

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