Task Routing on OlympusBench single-task setting
94.75AccuracyOlympus
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Olympus2024.12 | 94.75 | 95.8 | 94.75 | 95.77 | |
| HuggingGPTBase LLM=GPT-4o2024.12 | 81.35 | 85.54 | 81.55 | 83.56 | |
| HuggingGPTBase LLM=GPT-4o mini2024.12 | 70.14 | 76.51 | 72.14 | 75.46 |