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

Instruction Tuning research benchmarks – Page 2 · SOTA2 Research

Instruction Tuning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

74MMLU Score

4

May 15, 2026

Alpaca 52K instruction LLaMA-7B (test)

32.9BBH

4

Feb 26, 2026

Alpaca-style 500 samples (held-out)

1.347Cross-Entropy Loss (Std)

2

May 21, 2026

Anthropic HH (test)

56.3Win Rate

2

Feb 26, 2026

Alpaca-style instruction-tuning 5K samples (held-out)

1.342Cross-Entropy Loss (Std)

1

May 21, 2026

Anthropic HH-RLHF (test)

—

—Primary metric

0

Feb 26, 2026

—

—Primary metric

0

Feb 26, 2026

Page 2 of 2

Popular datasets

Dolly Alpaca Flan Anthropic-HH AlpacaEval

Follow for updates

Join SOTA2 Contact