Author
Sarath Kavuru
Sarath Kavuru leads Anthroly, which builds RL environments and datasets for AI labs. Research notes and papers on reward hacking, RL environments, and post-training data.
Profiles
Articles
- Where Should Step-Level Rewards for AI Agents Come From? Human Labels, Forked Rollouts, and Learned Verifiers
- How Many Expert Labels Does It Take to Fix Noisy Preference Data?
- How Do You Stop Reward Hacking in Rubric-Based RL?
- What Makes an RL Environment Training-Grade? Lessons from HF's RL Hub and TermGrade