Language Modeling on ProofPile 100K
3.19PerplexitySHAREDLLM
Evaluation Results
| Method | Links | |
|---|---|---|
| SHAREDLLMBase Model=LLaMA-2, Supervised fine-tuning=true2026.03 | 3.19 | |
| Activation BeaconBase Model=LLaMA-2, Supervised fine-tuning=true2026.03 | 3.31 | |
| StreamingLLMBase Model=LLaMA-2, Supervised fine-tuning=true2026.03 | 3.55 | |
| Activation BeaconBase Model=Mistral-7B, Supervised fine-tuning=true2026.03 | 3.72 | |
| SHAREDLLMBase Model=Mistral-7B, Supervised fine-tuning=true2026.03 | 3.74 | |
| StreamingLLMBase Model=Mistral-7B, Supervised fine-tuning=true2026.03 | 4.23 |