Scholar
Ted Moskovitz
Google Scholar ID: pPVXrTYAAAAJ
Anthropic
Citations & Impact
Citations
568
H-index
11
i10-index
14
Publications
20
Co-authors
0
Resume
Academic Achievements
- Confronting Reward Model Overoptimization with Constrained RLHF, ICLR 2024 (Spotlight, Top 5% of Submissions)
- ReLOAD: Reinforcement Learning with Optimistic Ascent-Descent for Last-Iterate Convergence in Constrained MDPs, ICML 2023
- Towards an Understanding of Default Policies in Multitask Policy Optimization, AISTATS 2022 (Best Paper Award Honorable Mention)
- A First-Occupancy Representation for Reinforcement Learning, ICLR 2022
- Tactical Optimism and Pessimism for Deep Reinforcement Learning, NeurIPS 2021
Research Experience
- Member of technical staff at Anthropic, working on scaling reinforcement learning; interned at DeepMind, worked on constrained reinforcement learning; interned at Uber AI Labs, worked on optimization for large-scale deep learning.
Education
- PhD: Gatsby Computational Neuroscience Unit (London), Advisors: Maneesh Sahani and Matt Botvinick
Background
- Research Interests: Building helpful intelligence with a grounded understanding of the world, particularly in reasoning, sequential decision-making, and optimization for large-scale models.
Miscellany
- Sometimes puts notes, code, and other writing online.