AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Before Thinking, Learn to Decide: Proactive Routing for Efficient Visual Reasoning

Yinan ZHOU, Haokun Lin, Yichen Wu, Yuxin Chen, Teng Wang, Caifeng Shan, Zhenan Sun, Chen Ma, Li Zhu, Ying Shan

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
use JRL routing instead of random routing at 60% to draftQwen2.5-VL-7B + Skywork-R1V-38B · MathVistaaccuracy: 71.6% -> 74.2%Improved
use JRL routing instead of token-probability routingQwen2.5-VL-7B + Skywork-R1V-38B · MathVerse (50% to draft)accuracy: 56.9% -> 58.5%Improved
use JRL routing instead of special-token routingQwen2.5-VL-7B + Skywork-R1V-38B · MathVista (60% to draft)accuracy: 72.1% -> 74.2%Improved
use DRL (draft rating only) routingQwen2.5-VL-7B draft · MathVerserouting AUC (%): 56.13 -> 55.44Hurt
add target rating (JRL) on top of DRL (ablation)Draft model · MathVerserouting AUC (%): 55.44 -> 56.97Improved
add target rating (JRL) on top of DRL (ablation)Draft model · ChartQArouting AUC (%): 81.97 -> 83.61Improved
add target rating (JRL) on top of DRL (ablation)Draft model · MathVistarouting AUC (%): 73.29 -> 73.36No clear effect
add Internal Probability Ranking within score bands (ablation)DRL/JRL routing · ChartQA and MathVistarouting qualityImproved
replace SFT-based rating training with RL-based rating training (ablation)Draft model · MathVistadraft accuracy: drops by 4.2% with SFTImproved
raise rating loss weight from 1.5 to 5 or 10 (ablation)DRL/JRL training · MathVerseaccuracyHurt
use larger tau for JRL score (ablation)JRL routing · MathVistarouting curveMixed
use DRL-trained draft for routing on ChartQA with target 84.39DRL · ChartQA (p_t>40)accuracy: 84.39 -> 84.72Improved

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