BigBench
Beyond the Imitation Game Benchmark
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to probe large language models and extrapolate their future capabilities.
How to run
ltb · bigbench
$ npm i -g lettokenburn-cli$ ltb bench pull bigbench --n 50$ ltb bench run ./bigbench -a claude-code -m anthropic/claude-opus-4-8$ ltb bench push
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
Scores
7 sourced entries| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| Chinchilla-70B | 92.8 | N/A | Independent | medium | Mar 29, 2022 | arxiv.org |
| Gopher-280B | 88.7 | N/A | Independent | medium | Dec 8, 2021 | arxiv.org |
| Gopher-280B | 77.5 | N/A | Independent | medium | Dec 8, 2021 | arxiv.org |
| Gopher-280B | 77.5 | N/A | Independent | medium | Dec 8, 2021 | arxiv.org |
| Chinchilla-70B | 77.4 | N/A | Independent | medium | Mar 29, 2022 | arxiv.org |
| Gopher-280B | 74.5 | N/A | Independent | medium | Dec 8, 2021 | arxiv.org |
| Gopher-280B | 28.2 | N/A | Independent | medium | Dec 8, 2021 | arxiv.org |
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Reproduced by the community
All runsLoading community runs…
About
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to probe large language models and extrapolate their future capabilities.
Available in EleutherAI lm-evaluation-harness as task family bigbench (286 tasks).