Stubbi
Rankings
Benchmark

BIG-bench

Probe language models on more than 200 crowd-contributed tasks.

Checked Oct 22, 2026BIG-bench collaboration

Explore the original evidence

Explore the original evaluation, its methods, and reported findings.

Open eval

Coverage

204 tasks contributed by more than 450 authors, spanning linguistics, math, commonsense, social bias, and more.

Methodology

Each task defines its own metric. BIG-Bench Lite and BIG-Bench Hard are commonly used subsets.

People & organizations

3

Sources

Original paper and author team