The Information Geometry of Unsupervised Reinforcement Learning

• URL: http://arxiv.org/abs/2110.02719v1
• Date: Wed, 6 Oct 2021 13:08:36 GMT
• Title: The Information Geometry of Unsupervised Reinforcement Learning
• Authors: Benjamin Eysenbach, Ruslan Salakhutdinov, and Sergey Levine
• Abstract summary: Unsupervised skill discovery is a class of algorithms that learn a set of policies without access to a reward function. We show that unsupervised skill discovery algorithms do not learn skills that are optimal for every possible reward function.
• Score: 133.20816939521941
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: How can a reinforcement learning (RL) agent prepare to solve downstream tasks if those tasks are not known a priori? One approach is unsupervised skill discovery, a class of algorithms that learn a set of policies without access to a reward function. Such algorithms bear a close resemblance to representation learning algorithms (e.g., contrastive learning) in supervised learning, in that both are pretraining algorithms that maximize some approximation to a mutual information objective. While prior work has shown that the set of skills learned by such methods can accelerate downstream RL tasks, prior work offers little analysis into whether these skill learning algorithms are optimal, or even what notion of optimality would be appropriate to apply to them. In this work, we show that unsupervised skill discovery algorithms based on mutual information maximization do not learn skills that are optimal for every possible reward function. However, we show that the distribution over skills provides an optimal initialization minimizing regret against adversarially-chosen reward functions, assuming a certain type of adaptation procedure. Our analysis also provides a geometric perspective on these skill learning methods.

Related papers

• Representation Learning with Multi-Step Inverse Kinematics: An Efficient and Optimal Approach to Rich-Observation RL [106.82295532402335]
  Existing reinforcement learning algorithms suffer from computational intractability, strong statistical assumptions, and suboptimal sample complexity. We provide the first computationally efficient algorithm that attains rate-optimal sample complexity with respect to the desired accuracy level. Our algorithm, MusIK, combines systematic exploration with representation learning based on multi-step inverse kinematics.
  arXiv  Detail & Related papers  (2023-04-12T14:51:47Z)
• Inapplicable Actions Learning for Knowledge Transfer in Reinforcement Learning [3.194414753332705]
  We show that learning inapplicable actions greatly improves the sample efficiency of RL algorithms. Thanks to the transferability of the knowledge acquired, it can be reused in other tasks and domains to make the learning process more efficient.
  arXiv  Detail & Related papers  (2022-11-28T17:45:39Z)
• A Generalist Neural Algorithmic Learner [18.425083543441776]
  We build a single graph neural network processor capable of learning to execute a wide range of algorithms. We show that it is possible to effectively learn algorithms in a multi-task manner, so long as we can learn to execute them well in a single-task regime.
  arXiv  Detail & Related papers  (2022-09-22T16:41:33Z)
• Towards Diverse Evaluation of Class Incremental Learning: A Representation Learning Perspective [67.45111837188685]
  Class incremental learning (CIL) algorithms aim to continually learn new object classes from incrementally arriving data. We experimentally analyze neural network models trained by CIL algorithms using various evaluation protocols in representation learning.
  arXiv  Detail & Related papers  (2022-06-16T11:44:11Z)
• MURAL: Meta-Learning Uncertainty-Aware Rewards for Outcome-Driven Reinforcement Learning [65.52675802289775]
  We show that an uncertainty aware classifier can solve challenging reinforcement learning problems. We propose a novel method for computing the normalized maximum likelihood (NML) distribution. We show that the resulting algorithm has a number of intriguing connections to both count-based exploration methods and prior algorithms for learning reward functions.
  arXiv  Detail & Related papers  (2021-07-15T08:19:57Z)
• Evolving Reinforcement Learning Algorithms [186.62294652057062]
  We propose a method for meta-learning reinforcement learning algorithms. The learned algorithms are domain-agnostic and can generalize to new environments not seen during training. We highlight two learned algorithms which obtain good generalization performance over other classical control tasks, gridworld type tasks, and Atari games.
  arXiv  Detail & Related papers  (2021-01-08T18:55:07Z)
• Towards Understanding the Behaviors of Optimal Deep Active Learning Algorithms [19.65665942630067]
  Active learning (AL) algorithms may achieve better performance with fewer data because the model guides the data selection process. There is little study on what the optimal AL looks like, which would help researchers understand where their models fall short. We present a simulated annealing algorithm to search for this optimal oracle and analyze it for several tasks.
  arXiv  Detail & Related papers  (2020-12-29T22:56:42Z)
• Discovering Reinforcement Learning Algorithms [53.72358280495428]
  Reinforcement learning algorithms update an agent's parameters according to one of several possible rules. This paper introduces a new meta-learning approach that discovers an entire update rule. It includes both 'what to predict' (e.g. value functions) and 'how to learn from it' by interacting with a set of environments.
  arXiv  Detail & Related papers  (2020-07-17T07:38:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.