pith. sign in

Amanda Askell

Identifiers

  • name variant Amanda Askell 0.60 · backfill

Papers (18)

  1. Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming cs.CL · 2025 · author #11
  2. Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training cs.CR · 2024 · author #12
  3. Towards Understanding Sycophancy in Language Models cs.CL · 2023 · author #5
  4. Towards Measuring the Representation of Subjective Global Opinions in Language Models cs.CL · 2023 · author #5
  5. Discovering Language Model Behaviors with Model-Written Evaluations cs.CL · 2022 · author #57
  6. Constitutional AI: Harmlessness from AI Feedback cs.CL · 2022 · author #4
  7. Measuring Progress on Scalable Oversight for Large Language Models cs.HC · 2022 · author #8
  8. In-context Learning and Induction Heads cs.LG · 2022 · author #8
  9. Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned cs.CL · 2022 · author #4
  10. Language Models (Mostly) Know What They Know cs.CL · 2022 · author #3
  11. Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models cs.CL · 2022 · author #24
  12. Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback cs.CL · 2022 · author #4
  13. Training language models to follow instructions with human feedback cs.CL · 2022 · author #16
  14. A General Language Assistant as a Laboratory for Alignment cs.CL · 2021 · author #1
  15. Learning Transferable Visual Models From Natural Language Supervision cs.CV · 2021 · author #8
  16. Language Models are Few-Shot Learners cs.CL · 2020 · author #10
  17. Release Strategies and the Social Impacts of Language Models cs.CL · 2019 · author #4
  18. The Role of Cooperation in Responsible AI Development cs.CY · 2019 · author #1

Mentions

  • 2501.18837 #11 · arxiv_oai · confidence 0.70 Amanda Askell
  • 2211.03540 #8 · arxiv_oai · confidence 0.70 Amanda Askell
  • 1908.09203 #4 · arxiv_oai · confidence 0.70 Amanda Askell
  • 2306.16388 #5 · arxiv_oai · confidence 0.70 Amanda Askell

Frequent Coauthors