Leon Derczynski
leonderczynski.bsky.social
LLMs & Security at NVIDIA Prof in CS/NLP at IT University of Copenhagen garak guy, garak.ai "berømt skikkelse" "like a gazelle" Seattle/Copenhagen 🏔️
👤 Joined on 20 November 2024
835
Followers
371
Following
328
Total Posts
📜 Kullanıcı Geçmişi
- 20 Kas 2024, 14:48
🎉 Bluesky'e katıldı - İlk kullanıcı adı:
@leonderczynski.bsky.social
🚫 Blokladığı Kullanıcılar
❤️ Beğendiği Gönderiler
Anchor your doc at the sloppified median, and it'll never move too far from it. Also: LLM slop is zero effort, and that 'zero effort' signal becomes part of the message of the piece. "Why should I read/edit/review what you couldn't be bothered to write?" &c.
Show PostNew NVIDIA AIRT research! Tamas Voros did great work adapting refusal direction to preventing indirect prompt injection. Really neat mech-interp approach that's very effective at stopping single-turn indirect prompt injection with minimal impact on normal use. research.nvidia.com/ai-security/...
Show Post
Hey, that only really matters if you have access to the hardware. What are they going to do put one out in public on every street corner for anyone to steal?
Show Post
Seems it's hackable enough to invalidate evidence?
Show Post