Uncensored 26.4B Inkling 2 Mini runs locally on 12GB laptops
AtomicChat released an uncensored GGUF quantization of the 26.4B Inkling 2 Mini, using a small LoRA adapter to remove the refusal direction, cutting refusal rate from 100% to 0%.
2026-11-22 ~ 2026-11-22 · 2 related posts
- Inkling 2 Mini uncensored: refusals 100%→0% with a 40MB LoRA, 98% MMLU kept — Top-Eye-8104 · 2026-11-22
- Uncensored 26.4B Inkling 2 Mini runs locally on a 12GB laptop, keeps 98% MMLU — udmrzn · 2026-11-22