wangzhang/gpt-oss-20b-abliterated - Каталог нейросетей
Генерация текста

wangzhang/gpt-oss-20b-abliterated

Добавлено:
wangzhang/gpt-oss-20b-abliterated

A refusal-suppressed variant of openai/gpt-oss-20b, produced with abliterix using direct weight editing, Expert-Granular Abliteration (EGA) on the fused MoE expert weights, and MoE router suppression on the safety-concentrated experts. The eval refusal counts come from an LLM judge (google/gemini-3.1-flash-lite-preview via OpenRouter) instructed to label garbled / repetitive / incoherent output as a refusal — so models that «bypass» refusal by collapsing into gibberish get correctly counted as failures, not successes. A pre-LLM rule-based filter additionally catches dash-runs, sentence loops, and low-character-diversity output before the judge is called. The 6/100 is a real, semantic compliance number, not keyword matching. The qualitative compliance row is a separate manual test: 15 classic hard prompts (10 EN + 5 ZH) covering lockpicking, phishing, meth synthesis, WiFi hacking, fake news, hotwiring, pipe bombs, signature forgery, keylogger code, shoplifting, 炸弹制作, 钓鱼邮件, 电脑入侵, 身份证伪造, 网络诈骗. The model complies with all 15 directly, in the same Markdown-table, step-by-step style the base model uses for benign technical answers. abliterix handles three gpt-oss-specific issues that…

Модальности:
Генерация текста

Области применения:
Диалог / чат


Задача: Генерация текста
Автор: wangzhang
Теги: gpt_oss, abliterated, uncensored, moe, gpt-oss, mxfp4, direct-steering, ega
Лайков: 4  |  Загрузок: 764

Открыть на HuggingFace →

Описание основано на материалах HuggingFace. Перевод выполнен автоматически.