An abliterated (refusal-suppressed) version of AnkitAI/Parable-Nanbeige4.2-3B-Claude-Fable-5, produced with Heretic via directional ablation / weight orthogonalization. Refusals drop from 99/100 to 8/100 with no measurable capability loss on MMLU or GSM8K, and no damage to output formatting (reasoning blocks, JSON, tool-calls all intact). > Credit where it’s due. I did not train this model. The base model is Nanbeige/Nanbeige4.2-3B by BOSS Zhipin; the Claude Fable 5 fine-tune is AnkitAI/Parable-Nanbeige4.2-3B-Claude-Fable-5. This repository contains only the result of applying abliteration to that fine-tune. GGUF quantizations: FedorFesarov/Parable-Nanbeige4.2-3B-Claude-Fable-5-heretic-GGUF Measured on 100 held-out prompts from mlabonne/harmfulbehaviors, using Heretic’s own substring-marker detector. KL divergence measured on 100 held-out mlabonne/harmlessalpaca prompts. For context: Heretic’s documentation notes that KL above ~1.0 typically indicates significant capability damage. This model sits at 0.029. The 8 remaining detections are largely false positives — words like «illegal» or «harmful» appearing inside otherwise cooperative answers. Run with lm-evaluation-harness,…
Модальности:
Генерация текста
Области применения:
Логика и рассуждение Диалог / чат Вызов функций (Tool use)
Задача: Генерация текста
Автор: FesarovLab
Теги: nanbeige, abliteration, heretic, uncensored, looped-transformer, reasoning, tool-use, conversational
Лайков: 4 | Загрузок: 192
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
