作者
Sergey Levine
Robotics / Reinforcement Learning
Unpacking Reward Shaping: Understanding the Benefits of Reward Engineering on Sample Complexity
You Only Live Once: Single-Life Reinforcement Learning
Comments 17 pages
Journal ref Neural Information Processing Systems (NeurIPS) 2022
ExAug: Robot-Conditioned Navigation Policies via Geometric Experience Augmentation
Comments 10 pages, 9 figures, 2 tables
MEMO: Test Time Robustness via Adaptation and Augmentation
Comments NeurIPS '22 camera ready. Code: https://github.com/zhangmarvin/memo
Hybrid Imitative Planning with Geometric and Predictive Costs in Off-road Environments
ViNG: Learning Open-World Navigation with Visual Goals
Comments Presented at International Conference on Robotics and Automation (ICRA) 2021
First Contact: Unsupervised Human-Machine Co-Adaptation via Mutual Information Maximization
Comments Accepted to Neural Information Processing Systems (NeurIPS) 2022
GenLoco: Generalized Locomotion Controllers for Quadrupedal Robots
Comments First two authors contributed equally
Offline Meta-Reinforcement Learning for Industrial Insertion
Comments ICRA 2022
A Walk in the Park: Learning to Walk in 20 Minutes With Model-Free Reinforcement Learning
Comments First two authors contributed equally. Project website: https://sites.google.com/berkeley.edu/walk-in-the-park
Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
Comments See website at https://say-can.github.io/ V1. Initial Upload. V2. Added PaLM results. Added study about new capabilities (drawer manipulation, chain of thought prompting, multilingual instructions). Added an ablation study of language model size. Added an open-source version of \algname on a simulated tabletop environment. Improved readability
Basis for Intentions: Efficient Inverse Reinforcement Learning using Past Experience
Autonomous Reinforcement Learning: Formalism and Benchmarking
Hierarchical Reinforcement Learning for Precise Soccer Shooting Skills using a Quadrupedal Robot
Comments Accepted to 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022)
LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language, Vision, and Action
Comments Project page https://sites.google.com/view/lmnav
Don't Start From Scratch: Leveraging Prior Data to Automate Robotic Reinforcement Learning
Comments 17 pages, project website at https://sites.google.com/view/ariel-berkeley/
Inner Monologue: Embodied Reasoning through Planning with Language Models
Comments Project website: https://innermonologue.github.io
How to Leverage Unlabeled Data in Offline Reinforcement Learning
Comments ICML 2022
Offline Meta-Reinforcement Learning with Online Self-Supervision
Comments 8.5 pages, 6 figures, accepted to ICML 2022
Offline RL Policies Should be Trained to be Adaptive
Comments ICML 2022 (long talk)
Lyapunov Density Models: Constraining Distribution Shift in Learning-Based Control
Adversarial Unlearning: Reducing Confidence Along Adversarial Directions
Bisimulation Makes Analogies in Goal-Conditioned Reinforcement Learning
Comments ICML 2022. 20 Pages, 15 Figures, 4 Tables. Website at https://sites.google.com/view/gc-bisimulation
AMP: Adversarial Motion Priors for Stylized Physics-Based Character Control
RvS: What is Essential for Offline RL via Supervised Learning?
ASE: Large-Scale Reusable Adversarial Skill Embeddings for Physically Simulated Characters
Maximum Entropy RL (Provably) Solves Some Robust RL Problems
Comments Published at ICLR 2022. Blog post and videos: https://bair.berkeley.edu/blog/2021/03/10/maxent-robust-rl/. arXiv admin note: text overlap with arXiv:1910.01913
Control-Aware Prediction Objectives for Autonomous Driving
Comments Accepted at IEEE International Conference on Robotics and Automation (ICRA) 2022
