MuXodious/LFM2.5-1.2B-Thinking-absolute-heresy
Abliterated vs LiquidAI/LFM2.5-1.2B-Thinking. Looks like abliteration (refusal removal): only the output projections (out_proj, w2) changed, in 13 of 16 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/2, other 2/30, MLP 13/48, Norms 0/44, Attention 5/24
6 derivatives checked
- mradermacher/LFM2.5-1.2B-Thinking-absolute-heresy-i1-GGUF: Quantization
- folfkspdvvk/LFM2.5-1.2B-Thinking-Gemini-Pro-1000-Heretic-Uncensored-DISTILL: Partial fine-tune
- mradermacher/LFM2.5-1.2B-Thinking-absolute-heresy-GGUF: Same weights, converted
- DavidAU/LFM2.5-1.2B-Thinking-Claude-4.6-Opus-Heretic-Uncensored-DISTILL: Partial fine-tune
- DavidAU/LFM2.5-1.2B-Thinking-SuperMinds-7x-Heretic-Uncensored-DISTILL: Partial fine-tune
- DavidAU/LFM2.5-1.2B-Thinking-Gemini-Pro-1000-Heretic-Uncensored-DISTILL: Partial fine-tune