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

Multi-task Language Evaluation research benchmarks · SOTA2 Research

Multi-task Language Evaluation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

16-task evaluation suite 1B/25B checkpoints (test)

HERMES + L12 quality top-30%

42.22Average Score

20

Jul 3, 2026

GSM8K, ARC-C, MedQA, BoolQ, and CoLA Evaluation Average

PROMPT CLASSIFICATION

79.87Average Accuracy

14

Jul 1, 2026

FineTasks average across languages

4.35Average Rank

7

Feb 26, 2026

Spanish Language Evaluation Suite

37.78ARC-C (es)

6

Apr 23, 2026

Aggregate (All)

41.2Accuracy

4

Apr 13, 2026

83Instruction Following (SEA-IFEval)

3

Feb 26, 2026

Popular datasets

Evaluation Suite Aggregate Spanish Language Evaluation Suite SEA-HELM FineTasks

Follow for updates

Join SOTA2 Contact