Benjamin Eysenbach, Ruslan Salakhutdinov, Sergey Levine

Research output: Contribution to conferencePaperpeer-review

14 Scopus citations


We study the problem of predicting and controlling the future state distribution of an autonomous agent. This problem, which can be viewed as a reframing of goal-conditioned reinforcement learning (RL), is centered around learning a conditional probability density function over future states. Instead of directly estimating this density function, we indirectly estimate this density function by training a classifier to predict whether an observation comes from the future. Via Bayes' rule, predictions from our classifier can be transformed into predictions over future states. Importantly, an off-policy variant of our algorithm allows us to predict the future state distribution of a new policy, without collecting new experience. This variant allows us to optimize functionals of a policy's future state distribution, such as the density of reaching a particular goal state. While conceptually similar to Q-learning, our work lays a principled foundation for goal-conditioned RL as density estimation, providing justification for goal-conditioned methods used in prior work. This foundation makes hypotheses about Q-learning, including the optimal goal-sampling ratio, which we confirm experimentally. Moreover, our proposed method is competitive with prior goal-conditioned RL methods.

Original languageEnglish (US)
StatePublished - 2021
Externally publishedYes
Event9th International Conference on Learning Representations, ICLR 2021 - Virtual, Online
Duration: May 3 2021May 7 2021


Conference9th International Conference on Learning Representations, ICLR 2021
CityVirtual, Online

All Science Journal Classification (ASJC) codes

  • Language and Linguistics
  • Computer Science Applications
  • Education
  • Linguistics and Language


Dive into the research topics of 'C-LEARNING: LEARNING TO ACHIEVE GOALS VIA RECURSIVE CLASSIFICATION'. Together they form a unique fingerprint.

Cite this