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

Home/Benchmarks

Compositional Reasoning on CLEVR-M

81.7Accuracy

MetaForge

Updated 1d ago

Evaluation Results

Method	Links
MetaForge 2026.06		81.7
MetaForge 2026.06		80.4

SOTA Paper

MetaForge

MetaForge: A Self-Evolving Multimodal Agent that Retrieves, Adapts, and Forges Tools On Demand

Dataset

Clevr

Follow for update

@wizwand_team Discord

Related Benchmarks

Tool Use on VerlTool IID Tools Tool Use on VerlTool OOD Tools Robust Visual Question Answering on RoBUT Visual Question Answering on Localized Narratives Webpage Understanding on WebSight

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs