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

Question Answering and Reasoning research benchmarks · SOTA2 Research

Question Answering and Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

ARC Mean of Easy and Challenge

94.75Accuracy

20

Jun 16, 2026

NQ, TriviaQA, HotpotQA, and GSM8K Average

72.5Average Accuracy

16

Jun 4, 2026

Downstream Reasoning Suite (Arc-e, PIQA, Hellaswag, OpenBookQA, Winogrande, MMLU, BoolQ)

49.23ARC-e

14

Feb 26, 2026

Macro-average (MMLU, MATH, GSM8K, BBH)

46Cost Reduction

8

Feb 27, 2026

Macro-average (MMLU, MATH, GSM8K, BBH) (test)

—

—Cost Reduction

0

Feb 26, 2026

Popular datasets

ARC Macro-average (MMLU, MATH, GSM8K, BBH)Macro Average Downstream Reasoning Suite NQ, TriviaQA, HotpotQA, and GSM8K

Follow for updates

Join SOTA2 Contact