primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
| model-00003-of-00006.safetensors | 4.61 GB | 78e1237c | download |
| model-00004-of-00006.safetensors | 4.61 GB | ca8c894b | download |
| model-00005-of-00006.safetensors | 4.61 GB | 3bd85fbd | download |
| model-00002-of-00006.safetensors | 4.61 GB | 3c66dc70 | download |
| model-00001-of-00006.safetensors | 4.57 GB | 320cbb72 | download |
| model-00006-of-00006.safetensors | 4.48 GB | 6003446a | download |
| tokenizer.json | 10.9 MB | 2c14cb8c | download |
| model.safetensors.index.json | 46.4 KB | 1ce7d1c6 | download |
| chat_template.jinja | 2.19 KB | c2066bd7 | download |
| config.json | 1.77 KB | 706525bb | download |
| .gitattributes | 1.53 KB | 52373fe2 | download |
| tokenizer_config.json | 421 B | 699cc3a3 | download |
| README.md | 276 B | 141e478c | download |
| generation_config.json | 181 B | ad8d4a0e | download |
Experimental modified / abliterated versions
Model: DeepSeek-R1-Distill-Qwen-14B
Trials: [200]
Status: Tested and saved
Original model: DeepSeek-R1-Distill-Qwen
License: [Mit]
Research and experimentation by [3MPER0RR]
Every model page has an "Open in app" button that hands off directly to a local runtime of your choice - Infrahuman, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.