robots:
E.S.T.H.E.R: an exciting wheelchair tennis robot with tremendous capabilities in both perception and action
Firebot: the first robot i worked with at UT that has capabilities to allow it to provide useful information for firefighters during fires
The Mirokaï: an emotive robot that can maneuver in tight spaces that i interacted with at SXSW 2023
Intuitive DaVinci: a precise surgical robot family whose capabilities to make medicine safer excite me
AirHockey UR5e: the robot and air hockey setup that serves as my research home
The Building-wide-intelligence bot: the first robot i ever worked with on a formal research project
games:
Poker: one of the most exciting imperfect information games explored by several papers
Stratego: as seen in a recent paper by DeepMind, this is a really fun imperfect information strategy game
Catan: a classic game that my friend group and i have become obsessed with. i hope to build an agent that can beat my friends in the near future :)
Ludo: a classic indian board game that has roots in ancient indian literature
papers:
(contemporary)
| Achieving Human Level Competitive Robot Table Tennis | 2024 |
| Direct Preference Optimization: Your Language Model is Secretly a Reward Model | 2023 |
| Diffusion Policy: Visuomotor Policy Learning via Action Diffusion | 2023 |
| Unsupervised Skill Discovery with Bottleneck Option Learning | 2021 |
| Reward is Enough | 2021 |
| Diversity is All You Need: Learning Skills Without a Reward Function | 2019 |
(older)
| Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor | 2018 |
| Learning Multi-Level Hierarchies with Hindsight | 2017 |
| Playing Atari with Deep Reinforcement Learning | 2013 |
| On the Complexity of Solving Markov Decision Problems | 1995 |
| Neural Networks and Physical Systems with Emergent Collective Computational Abilities | 1982 |
books: