Sotabase

Career

· Research Scientist, Anthropic2023–
· Researcher, Anthropic2022–
· Associate Professor of Data Science and Computer Science, New York University2016–
· PhD in Linguistics and Computer Science, Stanford University2011–2016
· Undergraduate Degree in Linguistics, University of Chicago2007–2011

Publications (109)

Constitutional AI: Harmlessness from AI Feedback
arXiv.org · 2022
2,410
cited
Language Models (Mostly) Know What They Know
arXiv.org · 2022
1,204
cited
Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting
Neural Information Processing Systems · 2023
779
cited
Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned
arXiv.org · 2022
656
cited
BBQ: A hand-built bias benchmark for question answering
Findings · 2021
631
cited
Discovering Language Model Behaviors with Model-Written Evaluations
Annual Meeting of the Association for Computational Linguistics · 2022
621
cited
Universal Dependencies 2.1
2017
418
cited
Measuring Faithfulness in Chain-of-Thought Reasoning
arXiv.org · 2023
321
cited
Pretraining Language Models with Human Preferences
International Conference on Machine Learning · 2023
279
cited
Studying Large Language Model Generalization with Influence Functions
arXiv.org · 2023
276
cited
QuALITY: Question Answering with Long Input Texts, Yes!
North American Chapter of the Association for Computational Linguistics · 2021
197
cited
Reasoning Models Don't Always Say What They Think
arXiv.org · 2025
195
cited
The Capacity for Moral Self-Correction in Large Language Models
arXiv.org · 2023
195
cited
Inverse Scaling: When Bigger Isn't Better
Trans. Mach. Learn. Res. · 2023
185
cited
Measuring Progress on Scalable Oversight for Large Language Models
arXiv.org · 2022
177
cited
Eight Things to Know about Large Language Models
Critical AI · 2023
143
cited
Question Decomposition Improves the Faithfulness of Model-Generated Reasoning
arXiv.org · 2023
108
cited
Comparing Test Sets with Item Response Theory
Annual Meeting of the Association for Computational Linguistics · 2021
52
cited
Two Failures of Self-Consistency in the Multi-Step Reasoning of LLMs
Trans. Mach. Learn. Res. · 2023
51
cited
(QA)^2: Question Answering with Questionable Assumptions
Annual Meeting of the Association for Computational Linguistics · 2022
49
cited
Sotabase
Sam Bowman | Researcher Profile | Sotabase | Sotabase