Genie: Generative Interactive Environments

no code implementations • 23 Feb 2024 • Jake Bruce, Michael Dennis, Ashley Edwards, Jack Parker-Holder, Yuge Shi, Edward Hughes, Matthew Lai, Aditi Mavalankar, Richie Steigerwald, Chris Apps, Yusuf Aytar, Sarah Bechtle, Feryal Behbahani, Stephanie Chan, Nicolas Heess, Lucy Gonzalez, Simon Osindero, Sherjil Ozair, Scott Reed, Jingwei Zhang, Konrad Zolna, Jeff Clune, Nando de Freitas, Satinder Singh, Tim Rocktäschel

We introduce Genie, the first generative interactive environment trained in an unsupervised manner from unlabelled Internet videos.

Provably Efficient Model-based Policy Adaptation

no code implementations • ICML 2020 • Yuda Song, Aditi Mavalankar, Wen Sun, Sicun Gao

The high sample complexity of reinforcement learning challenges its use in practice.

Continuous Control Meta-Learning

Goal-conditioned Batch Reinforcement Learning for Rotation Invariant Locomotion

no code implementations • 17 Apr 2020 • Aditi Mavalankar

For the locomotion task, this translates to data collection using a policy learnt by the agent for walking straight in one direction, and using that data to learn a goal-conditioned policy that enables the agent to walk in any direction.

reinforcement-learning Reinforcement Learning (RL)

Learning Latent Representations for Inverse Dynamics using Generalized Experiences

no code implementations • 25 Sep 2019 • Aditi Mavalankar, Sicun Gao

Many practical robot locomotion tasks require agents to use control policies that can be parameterized by goals.

Navigate