AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Explicit Logic Channel for Validation and Enhancement of MLLMs on Zero-Shot Tasks

Mei Chee Leong, Gu Ying, Hui Tan, Liyuan Li, Nancy Chen

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Chain-of-thought multimodal reasoning · 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
add aligned fusion of ELC with the MLLMInternVL2.5-8B · NegBench COCOaccuracy: 0.912 -> 0.965Improved
add aligned fusion of ELC with the MLLMInternVL2.0-8B · NegBench COCOaccuracy: 0.4878 -> 0.8434Improved
add aligned fusion of ELC with the MLLMLLaVA-1.5-13B · NegBench COCOaccuracy: 0.6206 -> 0.6206No clear effect
add aligned fusion of ELC with the MLLMGemma-3-12B · HC-RefCOCOg testAcc@0.5: .088 -> .752Improved
add aligned fusion of ELC with the MLLMQwen3.0-VL-8B · HC-RefCOCOg testAcc@0.5: .818 -> .856Improved
add aligned fusion of ELC with the MLLMGemma-3-12B · HC-RefLoCo testAcc@0.5: .0920 -> .4270Improved
add aligned fusion of ELC with the MLLMQwen3.0-VL-8B · HC-RefLoCo testAcc@0.5: .9418 -> .9388No clear effect
use Qwen3.0 instead of Mistral as ELC language parser (ablation)ELC · NegBench COCOELC accuracy: 0.7577 -> 0.7614No clear effect
use Qwen3.0 instead of Mistral as ELC language parser (ablation)ELC · HC-RefCOCOg valELC Acc@0.5: 0.8074 -> 0.7996No clear effect
use Mistral instead of Qwen3.0 as ELC language parser (ablation)ELC · HC-RefLoCo valELC Acc@0.5: .8044 -> .8110No clear effect
mask one critical object in each image (adversarial) (ablation)Qwen3.0-VL ILC · HC-RefCOCOg adversarial (100 images)Acc@0.5: 91% -> 60%Hurt
mask one critical object in each image (adversarial) (ablation)ELC · HC-RefCOCOg adversarial (100 images)Acc@0.5: 82% -> 41%Hurt

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