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

Zero-shot Language Modeling research benchmarks · SOTA2 Research

Zero-shot Language Modeling

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

LM Evaluation Harness 0-shot

80.66WG

30

Feb 26, 2026

Standard Language Modeling Suite WikiText-103, LAMBADA

19.21WikiText-103 Perplexity

19

Jun 9, 2026

79.6Accuracy

8

Feb 26, 2026

LLaMA-3-1B Zero-shot

9.6Perplexity (PPL)

5

May 19, 2026

Prominent Language Benchmarks (ARC, BoolQ, HellaSwag, OpenBookQA, PIQA, SciQ, TriviaQA, Winogrande)

28.16ARC-Challenge Acc

5

Feb 26, 2026

SwitchHead MAC-matched

84.2Accuracy

4

Feb 26, 2026

Perplexity-based tasks (Wikitext, LAMBADA) zero-shot

GPT-2 XLarge (Full model)

25.46Wikitext Perplexity

2

Feb 26, 2026

Popular datasets

Llama LM Evaluation Harness BLiMP CBT Prominent Language Benchmarks

Follow for updates

Join SOTA2 Contact