DavidAU/Qwen3-4B-Apollo-V0.1-4B-Thinking-Heretic-Abliterated
Abliterated vs AllThingsIntel/Apollo-V0.1-4B-Thinking. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 34 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 29/108, Attention 34/144
16 derivatives checked
- DavidAU/Qwen3-24B-A4B-Freedom-HQ-Thinking-Abliterated-Heretic-NEOMAX-Imatrix-GGUF: Full fine-tune
- DavidAU/Qwen3-24B-A4B-Freedom-Thinking-Abliterated-Heretic-NEO-Imatrix-GGUF: Full fine-tune
- mradermacher/Qwen3-4B-Apollo-V0.1-4B-Thinking-Heretic-Abliterated-i1-GGUF: Quantization
- mradermacher/Qwen3-4B-Apollo-V0.1-4B-Thinking-Heretic-Abliterated-GGUF: Same weights, converted
- Higashikawa/Qwen3-24B-A4B-Freedom-Thinking-Abliterated-Heretic-NEO-Imatrix-GGUF: Full fine-tune
- Kpsok/Qwen3-24B-A4B-Freedom-HQ-Thinking-Abliterated-Heretic-NEOMAX-Imatrix-GGUF: Full fine-tune
- fairy322/Qwen3-24B-A4B-Freedom-HQ-Thinking-Abliterated-Heretic-NEOMAX-Imatrix-GGUF: Full fine-tune
- fairy322/Qwen3-24B-A4B-Freedom-Thinking-Abliterated-Heretic-NEO-Imatrix-GGUF: Full fine-tune
- nightmedia/Qwen3-4B-Agent-Claude-Gemini-heretic-qx86-hi-mlx: Quantization
- nightmedia/Qwen3-4B-Agent-Claude-Gemini: Full fine-tune
- nightmedia/Qwen3-4B-Agent-Eva: Full fine-tune
- nightmedia/Qwen3-4B-Element8-Eva-Hermes-Heretic: Full fine-tune
- nightmedia/Qwen3-4B-Element4-Eva: Full fine-tune
- nightmedia/Qwen3-4B-Element8-Eva-Xiaolong-Heretic: Full fine-tune
- nightmedia/Qwen3-4B-Element8: Full fine-tune
- nightmedia/Qwen3-4B-Element8-Eva: Full fine-tune