Zero-shot Commonsense Reasoning on Zero-shot tasks (test)
61Average AccuracyOriginal
Evaluation results
| Method | Links | |
|---|---|---|
| OriginalModel=Llama-30B2026.02 | 61 | |
| SVD-LLM v2Model=Llama-30B, Status=paper-claimed2026.02 | 60 | |
| COMPOTModel=Llama-30B2026.02 | 60 | |
| OriginalModel=Llama-13B2026.02 | 59 | |
| COMPOTModel=Llama-13B2026.02 | 57 | |
| SVD-LLMModel=Llama-30B2026.02 | 57 | |
| SVD-LLM v2Model=Llama-13B, Status=paper-claimed2026.02 | 56 | |
| SVD-LLMModel=Llama-13B2026.02 | 55 | |
| ASVDModel=Llama-13B2026.02 | 54 | |
| ASVDModel=Llama-30B2026.02 | 44 | |
| FWSVDModel=Llama-13B2026.02 | 43 | |
| FWSVDModel=Llama-30B2026.02 | 42 |