SciMIF Benchmark

What is SciMIF?

Evaluates instruction following of MLLMs across five scientific disciplines with a taxonomy of 10 constraint groups.

Released

2026-08-26

Evaluates

Materials & Chemistry, Multimodal

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2608.25973

SciMIF paper and code

• Paper https://arxiv.org/abs/2608.25973
• Code https://github.com/shenye7436/SciMIF

Which sources cite SciMIF?

• SciMIF

Related Materials & Chemistry, Multimodal benchmarks

• 3DCodeBench
• A-Bench
• A-OKVQA
• ActionAtlas
• ActiveFly-Bench
• Agent Planning Benchmark
• AgentHazard
• Agentic Evaluation in 2D Mazes