LLM Routing on SUPERGPQA (val)
0.776Top-1 AccORACLE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ORACLEModel Pool=Pool Large, Router Category=Oracle2026.01 | 0.776 | — | |
| ORACLEModel Pool=Pool Small, Router Category=Oracle2026.01 | 0.678 | — | |
| CASCAL-GTModel Pool=Pool Large, Router Category=Query-answer Routers2026.01 | 0.547 | 54.7 | |
| TOP-1 / TOP-3Model Pool=Pool Large, Router Category=Baseline2026.01 | 0.544 | 54.7 | |
| AVENGERSModel Pool=Pool Large, Router Category=Query-answer Routers2026.01 | 0.537 | — | |
| LLMRANKModel Pool=Pool Large, Router Category=Query-answer Routers2026.01 | 0.536 | — | |
| CASCALModel Pool=Pool Large, Router Category=Query-only Routers2026.01 | 0.534 | 53.8 | |
| SMOOTHIEModel Pool=Pool Large, Router Category=Query-only Routers2026.01 | 0.435 | 46.6 | |
| TOP-1 / TOP-3Model Pool=Pool Small, Router Category=Baseline2026.01 | 0.373 | 35.7 | |
| AVENGERSModel Pool=Pool Small, Router Category=Query-answer Routers2026.01 | 0.371 | — | |
| CASCAL-GTModel Pool=Pool Small, Router Category=Query-answer Routers2026.01 | 0.369 | 31 | |
| CASCALModel Pool=Pool Small, Router Category=Query-only Routers2026.01 | 0.324 | 30.8 | |
| LLMRANKModel Pool=Pool Small, Router Category=Query-answer Routers2026.01 | 0.318 | — | |
| SMOOTHIEModel Pool=Pool Small, Router Category=Query-only Routers2026.01 | 0.286 | 30.5 |