Timezone: »

Bisimulation Makes Analogies in Goal-Conditioned Reinforcement Learning
Philippe Hansen-Estruch · Amy Zhang · Ashvin Nair · Patrick Yin · Sergey Levine

Thu Jul 21 01:45 PM -- 01:50 PM (PDT) @ Room 309

Building generalizable goal-conditioned agents from rich observations is a key to reinforcement learning (RL) solving real world problems. Traditionally in goal-conditioned RL, an agent is provided with the exact goal they intend to reach. However, it is often not realistic to know the configuration of the goal before performing a task. A more scalable framework would allow us to provide the agent with an example of an analogous task, and have the agent then infer what the goal should be for its current state. We propose a new form of state abstraction called goal-conditioned bisimulation that captures functional equivariance, allowing for the reuse of skills to achieve new goals. We learn this representation using a metric form of this abstraction, and show its ability to generalize to new goals in real world manipulation tasks. Further, we prove that this learned representation is sufficient not only for goal-conditioned tasks, but is amenable to any downstream task described by a state-only reward function.

Author Information

Philippe Hansen-Estruch (University of California, Berkeley)

Undergraduate Researcher at BAIR, UC Berkeley - Working with Dr. Amy Zhang and Professor Sergey Levine

Amy Zhang (FAIR / UC Berkeley)
Ashvin Nair
Patrick Yin (UC Berkeley)
Sergey Levine (University of California, Berkeley)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors