Sid Reddy

Google Scholar | Github | Twitter | Resume
sgr [at] berkeley [dot] edu

I'm a second-year computer science Ph.D. student at the Berkeley Artificial Intelligence Research Lab, where I have the privilege of being co-advised by Sergey Levine and Anca Dragan. I'm supported by an NVIDIA Graduate Fellowship. This summer, I'll be interning at DeepMind, working with the AI safety team on reward modeling and machine theory of mind.

Research Interests

I'm currently exploring the intersection of machine learning, robotics, and cognitive science as a member of the RAIL and InterACT labs. I'm interested in combining human and machine intelligence to solve sequential decision-making problems that neither can on their own. My research statement contains details.

In my undergraduate research, I developed a mathematical framework for spaced repetition that can be used to improve long-term human learning, in collaboration with Thorsten Joachims, Siddhartha Banerjee, and Igor Labutov. Our work on the Leitner Queue Network combines ideas from queueing theory and psychology to optimize review scheduling in flashcard software. I also worked on the Latent Skill Embedding, a probabilistic model of student knowledge and educational content that can be used to recommend personalized lesson sequences in online courses.

Conference Papers

Siddharth Reddy, Anca D. Dragan, Sergey Levine, Where Do You Think You're Going?: Inferring Beliefs about Dynamics from Behavior, Neural Information Processing Systems, 2018.
[PDF] [Videos] [Code]

Siddharth Reddy, Anca D. Dragan, Sergey Levine, Shared Autonomy via Deep Reinforcement Learning, Robotics: Science and Systems, 2018.
[PDF] [Blog] [Videos] [Code]

Siddharth Reddy, Igor Labutov, Siddhartha Banerjee, Thorsten Joachims, Unbounded Human Learning: Optimal Scheduling for Spaced Repetition, ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016.
[PDF] [Code]

Workshop Papers

Siddharth Reddy, Anca D. Dragan, Sergey Levine, What Would pi* Do?: Imitation Learning via Off-Policy Reinforcement Learning, Deep Reinforcement Learning Workshop @ NeurIPS, 2018.
[PDF] [OpenReview]

Siddharth Reddy, Sergey Levine, Anca D. Dragan, Accelerating Human Learning with Deep Reinforcement Learning, NeurIPS Workshop on Teaching Machines, Robots, and Humans, 2017.
[PDF] [Code]

Siddharth Reddy, Igor Labutov, Thorsten Joachims, Latent Skill Embedding for Personalized Lesson Sequence Recommendation, ICML Workshop on Machine Learning for Education, 2015.
[PDF] [Code]