Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

General Visual Understanding on MME-RealWorld

63.8Score

DCRL

Updated 1mo ago

Evaluation Results

Method	Links
DCRL 2026.06		63.8
Qwen3-VL-8B 2026.06		62.8

SOTA Paper

DCRL

Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching

Dataset

MME-RealWorld

Follow for update

@wizwand_team Discord

Related Benchmarks

Wide-Baseline Correspondence Reasoning on ReasonMatch-Bench Overall 90-sample high-divergence subset Wide Baseline Matching on ReasonMatch-Bench Overall (test)Wide Baseline Matching on ReasonMatch-Bench Indoor (test)Wide Baseline Matching on ReasonMatch-Bench Outdoor (test)Wide Baseline Matching on ReasonMatch-Bench Object (test)

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs