Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Vision-Language Compositional Reasoning research benchmarks · SOTA2 Research

Vision-Language Compositional Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Winoground 1.0 (test)

89.5Text Score

23

May 22, 2026

66.24Accuracy

20

Apr 16, 2026

0.804Accuracy

14

Feb 26, 2026

ARO Benchmark Visual Genome Flickr30k MS-COCO (test)

89.3VG Attribution

11

Feb 27, 2026

44.2Relation Score

10

Jun 12, 2026

Winoground standard (test)

75.5Text Score

7

Feb 26, 2026

Winoground (test)

61.3Text Score

7

May 18, 2026

71.4VG-Rel

4

Feb 26, 2026

Winoground clean (no-tag)

32.16Text Score

2

Feb 26, 2026

Popular datasets

Winoground ARO SugarCrepe ARO Benchmark What's Up

Follow for updates

Join SOTA2 Contact