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

Common Sense Reasoning research benchmarks – Page 4 · SOTA2 Research

Common Sense Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Contextual Calibration (CC)

6.6ΔAcc

6

Feb 26, 2026

90.8EN Accuracy

6

Feb 26, 2026

0.34HellaSwag Delta

5

Jun 19, 2026

53.8Accuracy

5

Jun 4, 2026

HellaSwag topic-filtered

68.2Accuracy

5

Jun 4, 2026

ARC-c, ARC-e, HS, LAMB, OBQA, PIQA, SciQ, Wino

25.13ARC-c Accuracy

5

May 29, 2026

In-Domain Tasks ARC-e, ARC-c, BoolQ, HellaSwag, PIQA

74.7ARC-e Accuracy

5

May 21, 2026

64Accuracy

4

Feb 26, 2026

Winogrande (test)

77.9Score

4

Feb 26, 2026

ARC-Challenge In-Distribution Llama-3.1-8B (train test)

81.79Accuracy

3

Apr 7, 2026

64.5E*S

2

Jun 4, 2026

Common-sense Reasoning Benchmarks (BoolQ, SciQ, PIQA, WinoG., ARC-C, HellaS.) zero-shot

81BoolQ Accuracy

2

Jun 2, 2026

5.2CF Score

2

May 27, 2026

76.3Normalized Log Accuracy

2

Mar 18, 2026

Page 4 of 4

Popular datasets

ARC HellaSwag WinoGrande Common Sense Reasoning Tasks Common Sense Reasoning

Follow for updates

Join SOTA2 Contact