Agent Toolchain Scheduling on CCAD
55AccuracyLLMBOOST
Evaluation results
| Method | Links | |
|---|---|---|
| LLMBOOSTBase Model=Llama-3.1-8B Base, Params=2 × 8B2025.12 | 55 | |
| VOTEBase Model=Llama-3.1-8B Base, Params=2 × 8B2025.12 | 53 | |
| UNITEBase Model=Llama-3.1-8B Base, Params=2 × 8B2025.12 | 52.5 | |
| SingleBase Model=Llama-3.1-8B Base, Params=8B2025.12 | 52 | |
| LLMBOOSTBase Model=Qwen-2.5-7B Base, Params=2 × 7B2025.12 | 45.5 | |
| VOTEBase Model=Qwen-2.5-7B Base, Params=2 × 7B2025.12 | 43 | |
| UNITEBase Model=Qwen-2.5-7B Base, Params=2 × 7B2025.12 | 41.5 | |
| SingleBase Model=Qwen-2.5-7B Base, Params=7B2025.12 | 41 |