Goekdeniz-Guelmez/Josiefied-Qwen3-8B-abliterated-v1
Abliterated vs Qwen/Qwen3-8B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 35 of 36 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/145, MLP 35/108, Attention 35/144, lm_head 0/1
13 derivatives checked
- DavidAU/Qwen3-8B-64k-Context-2X-Josiefied-Uncensored: Identical copy
- bartowski/Goekdeniz-Guelmez_Josiefied-Qwen3-8B-abliterated-v1-GGUF: Same weights, converted
- mradermacher/Josiefied-Qwen3-8B-abliterated-v1-GGUF: Same weights, converted
- mlx-community/Josiefied-Qwen3-8B-abliterated-v1-4bit: Quantization
- mradermacher/Josiefied-Qwen3-8B-abliterated-v1-i1-GGUF: Quantization
- deadbydawn101/RavenX-Trade-8B-GGUF: LoRA-style
- FrostR/Josiefied-Qwen3-8B-abliterated-v1-Q4_K_M-GGUF: Quantization
- DevQuasar/Goekdeniz-Guelmez.Josiefied-Qwen3-8B-abliterated-v1-GGUF: Same weights, converted
- deadbydawn101/RavenX-Trade-8B-MAP-128k-mlx-4bit: LoRA-style
- RogerS-01/Josiefied-Qwen3-8B-abliterated-v1-AWQ: Quantization
- DavidAU/Qwen3-8B-192k-Context-6X-Josiefied-Uncensored: Identical copy
- lefantom00/Qwen3-8B-Josiefied-iSMART: LoRA-style
- richardbenson91477/Lisa-v0.2-8b: LoRA-style
3 other repos hold byte-for-byte identical weight files.