Sotabase

Career

· Member Of Technical Staff Research Scientist, OpenAI2025–
· PhD Artificial Intelligence Student, UC Berkeley2020–2025

Publications (27)

Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback
Trans. Mach. Learn. Res. · 2023
738
cited
On the Utility of Learning about Humans for Human-AI Coordination
Neural Information Processing Systems · 2019
488
cited
Humanity's Last Exam
Robotics · 2025
284
cited
Harms from Increasingly Agentic Algorithmic Systems
Conference on Fairness, Accountability and Transparency · 2023
198
cited
Engagement, user satisfaction, and the amplification of divisive content on social media
PNAS Nexus · 2023
73
cited
Characterizing Manipulation from AI Systems
Conference on Equity and Access in Algorithms, Mechanisms, and Optimization · 2023
65
cited
Estimating and Penalizing Induced Preference Shifts in Recommender Systems
International Conference on Machine Learning · 2022
49
cited
OpenAI GPT-5 System Card
2025
46
cited
AI Alignment with Changing and Influenceable Reward Functions
International Conference on Machine Learning · 2024
43
cited
Beyond Preferences in AI Alignment
Philosophical Studies · 2024
43
cited
Evaluating the Robustness of Collaborative Agents
Adaptive Agents and Multi-Agent Systems · 2021
41
cited
On Targeted Manipulation and Deception when Optimizing LLMs for User Feedback
International Conference on Learning Representations · 2024
41
cited
Estimating and Penalizing Preference Shift in Recommender Systems
ACM Conference on Recommender Systems · 2021
29
cited
UniMASK: Unified Inference in Sequential Decision Problems
Neural Information Processing Systems · 2022
27
cited
Twitter's Algorithm: Amplifying Anger, Animosity, and Affective Polarization
arXiv.org · 2023
18
cited
Optimal Behavior Prior: Data-Efficient Human Models for Improved Human-AI Collaboration
arXiv.org · 2022
15
cited
Towards Flexible Inference in Sequential Decision Problems via Bidirectional Transformers
arXiv.org · 2022
10
cited
Who Needs to Know? Minimal Knowledge for Optimal Coordination
International Conference on Machine Learning · 2023
8
cited
Time-Efficient Reward Learning via Visually Assisted Cluster Ranking
arXiv.org · 2022
7
cited
Defining Deception in Decision Making
Adaptive Agents and Multi-Agent Systems · 2024
4
cited
Sotabase
Micah Carroll | Researcher Profile | Sotabase | Sotabase