prithivMLmods/Qwen3.6-27B-abliterated-rMAX
Abliterated vs Qwen/Qwen3.6-27B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 64 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/4, MLP 64/304, Attention 64/556, lm_head 0/1, Norms 0/319
7 derivatives checked
- mradermacher/Qwen3.6-27B-abliterated-rMAX-i1-GGUF: Quantization
- mradermacher/Qwen3.6-27B-abliterated-rMAX-GGUF: Quantization
- prithivMLmods/Qwen3.6-27B-abliterated-rMAX-GGUF: Same weights, converted
- prithivMLmods/Qwen3.6-27B-abliterated-rMAX-Q4_K_M-GGUF: Quantization
- prithivMLmods/CapQwen3.6-27B-BLIP3o-Long-Caption-Distilled: Same weights, repacked
- prithivMLmods/Q3.6-27B-GLM-5.1-DA: LoRA-style
- prithivMLmods/Q3.6-27B-DS-v4-Flash-DA: LoRA-style