Apurva Badithela
Papers
1
Total Citations
1
H-Index
1
About
Apurva Badithela is a rising researcher at the intersection of robotics, imitation learning, and formal methods. Her work focuses on developing rigorous frameworks for evaluating and comparing learned robot policies, particularly in complex manipulation tasks where traditional metrics fall short. In her highly cited 2025 paper, "Is Your Imitation Learning Policy Better than Mine? Policy Comparison with Near-Optimal Stopping," she addresses a critical gap in robotics: how to fairly and efficiently determine when one learned policy outperforms another. By introducing a near-optimal stopping criterion for evaluation trials, Badithela provides a principled method for policy comparison that reduces wasteful experimentation while maintaining statistical rigor. This contribution is foundational for the growing field of dexterous manipulation, where long-horizon tasks demand reliable validation. Her work bridges formal verification and practical robotics, offering tools that enable researchers to benchmark imitation learning methods with confidence. With her innovative approach to policy evaluation, Badithela is shaping how the community assesses progress in robot learning, ensuring that new methods are not just proposed, but meaningfully compared.
Research Focus
Key Achievements
Top Papers
- 1