AssistantBench
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Reward Modeling | AssistantBench | Pairwise Accuracy89.17 | 13 | |
| Web Reasoning | AssistantBench | Accuracy31.8 | 8 | |
| Agentic Task Performance | AssistantBench (test) | Easy Accuracy65.8 | 6 |
Showing 3 of 3 rows