BIG-bench
diversePending Verification
Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark of 204 diverse tasks.
Published: 2022
Score Range: 0-100
Top Score: 93.1
BIG-bench Leaderboard
| Rank | Model | Provider | Score | Parameters | Released | Type |
|---|---|---|---|---|---|---|
| 1 | Claude 3.5 Sonnet | Anthropic | 93.1 | 2024-06-20 | Multimodal | |
| 2 | Claude 3 Opus | Anthropic | 86.8 | 2024-03-04 | Multimodal | |
| 3 | Claude 3 Sonnet | Anthropic | 82.9 | 2024-03-04 | Multimodal | |
| 4 | Claude 3 Haiku | Anthropic | 73.7 | 2024-03-04 | Multimodal | |
| 5 | Gemini Diffusion | 15 | 2025-05-20 | Text |
About BIG-bench
Methodology
BIG-bench evaluates model performance using a standardized scoring methodology. Scores are reported on a scale of 0 to 100, where higher scores indicate better performance. For detailed information about the scoring system and methodology, please refer to the original paper.
Publication
This benchmark was published in 2022.Read the full paper