API-based two-layer LLM cascade
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
API-based two-layer LLM cascade v1 (test)
93.5Cascade (Φ) Accuracy
5
API-based two-layer LLM cascade v1 (test)
94Cascade Accuracy (Φ)
2
Loading the SOTA2 catalog…