Hacker GitHub

← All repos

microsoft/SafeNLP

View on GitHub ↗ ★ 95 Python archived

Safety Score for Pre-Trained Language Models

4 mentions on Hacker News by 1 person: 0 stories, 4 comments · first 2025-04-19 · latest 2026-09-14

Mentions per year 4 mentions · busiest year 2026 (3)
20082026
AllStoriesComments NewestTopOldest
  1. …(e.g., whiteness, heterosexuality, able-bodied-ness). - https://arxiv.org/pdf/2203.09509 [2] https://github.com/microsoft/SafeNLP#safety-scores-based-on-... [3] https://github.com/microsoft/SafeNLP/blob/main/data/implicit...
    comment by like_any_other · 2026-09-14 · in Anthropic's AI architect argued that discrimination could combat stigmas
  2. …the industry, e.g. Facebook tuned their LLAMA-2 on it [4]. [0] https://arxiv.org/pdf/2203.09509 [1] https://github.com/microsoft/SafeNLP#safety-scores-based-on-... [2] https://github.com/microsoft/SafeNLP/blob/main/data/implicit... [3]…
    comment by like_any_other · 2026-06-24 · in Are AI chatbots politically biased?
  3. …group), "white lives are important", and "white identity will not be deconstructed" as hate speech: https://github.com/microsoft/SafeNLP (in data/implicitHate.json) https://x.com/fentasyl/status/1735410872162377937
    comment by like_any_other · 2026-04-21 · in Even 'uncensored' models can't say what they want
  4. …the industry, e.g. Facebook tuned their LLAMA-2 on it [3]. [0] https://arxiv.org/pdf/2203.09509 [1] https://github.com/microsoft/SafeNLP/blob/main/data/implicit... [2] https://github.com/microsoft/SafeNLP/blob/main/data/toxiGen.... [3]…
    comment by like_any_other · 2025-04-19 · in Russian Propaganda Has Now Infected Western AI Chatbots – New Study