Nova DasSarma
Source-listed: Systems Lead (Member of Technical Staff) · Anthropic
Record snapshot: Sep 3, 2026. Coverage and affiliations may be incomplete or historical.
Selected work
5- Training a Helpful and Harmless Assistant with Reinforcement Learning from Human FeedbackSaved credit: Contributor · 2022Explore people connected to this work →
- Constitutional AI: Harmlessness from AI FeedbackSaved credit: Contributor · 2022Explore people connected to this work →
- Language Models (Mostly) Know What They KnowSaved credit: Contributor · 2022Explore people connected to this work →
- Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons LearnedSaved credit: Contributor · 2022Explore people connected to this work →
- Sleeper Agents: Training Deceptive LLMs that Persist Through Safety TrainingSaved credit: Contributor · 2024Explore people connected to this work →