mlsanityData to 14 Oct 2026

Minimum Description Length Control

arXiv 2207.08258 ↗First posted 17 Jul 2022cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

We propose a novel framework for multitask reinforcement learning based on the minimum description length (MDL) principle. In this approach, which we term MDL-control (MDL-C), the agent learns the common structure among the tasks with which it is faced and then distills it into a simpler representation which facilitates faster convergence and generalization to new tasks. In doing so, MDL-C naturally balances adaptation to each task with epistemic uncertainty about the task distribution. We motivate MDL-C via formal connections between the MDL principle and Bayesian inference, derive theoretical performance guarantees, and demonstrate MDL-C's empirical effectiveness on both discrete and high-dimensional continuous control tasks.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 41 of its 78 references that are in the map. First posted on arXiv 17 Jul 2022.

Builds on

41 of 78 references.

References · 41

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  4. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  5. 2015Trust Region Policy Optimizationcited by 2,580
  6. 2021Convex optimizationcited by 2,515
  7. 2016Asynchronous Methods for Deep Reinforcement Learningcited by 2,258
  8. 2019Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learningcited by 1,204
  9. 2018DeepMind Control Suitecited by 956
  10. 2015Variational Dropout and the Local Reparameterization Trickcited by 490
  11. 2018Reinforcement Learning and Control as Probabilistic Inference: Tutorial and Reviewcited by 482
  12. 2019Online Learning: A Modern Introduction Using Convex Optimizationcited by 417
29 more
  1. 2018Maximum a Posteriori Policy Optimisationcited by 328
  2. 2017Variational Dropout Sparsifies Deep Neural Networkscited by 265
  3. 2017Distral: Robust Multitask Reinforcement Learningcited by 233
  4. 2018Distributed Distributional Deterministic Policy Gradientscited by 199
  5. 2019On the Theory of Policy Gradient Methods: Optimality, Approximation, and Distribution Shiftcited by 177
  6. 2017Equivalence Between Policy Gradients and Soft Q-Learningcited by 173
  7. 2022Evolving Curricula with Regret-Based Environment Designcited by 173
  8. 2017Bayesian Compression for Deep Learningcited by 116
  9. 2019InfoBot: Transfer and Exploration via the Information Bottleneckcited by 78
  10. 2018The Description Length of Deep Learning Modelscited by 72
  11. 2021On the Expressivity of Markov Rewardcited by 71
  12. 2013Sample Complexity of Multi-task Reinforcement Learningcited by 58
  13. 2019Information asymmetry in KL-regularized RLcited by 53
  14. 2021Same State, Different Task: Continual Reinforcement Learning without Interferencecited by 43
  15. 2020Behavior Priors for Efficient Reinforcement Learningcited by 41
  16. 2017Sparse Markov Decision Processes with Causal Sparse Tsallis Entropy Regularization for Reinforcement Learningcited by 37
  17. 2019Minimum Description Length Revisitedcited by 37
  18. 2019Learning to Score Behaviors for Guided Policy Optimizationcited by 34
  19. 2020Efficient Wasserstein Natural Gradients for Reinforcement Learningcited by 23
  20. 2018VIREL: A Variational Inference Framework for Reinforcement Learningcited by 22
  21. 2021Discovering a set of policies for the worst case rewardcited by 21
  22. 2020Predictive Complexity Priorscited by 14
  23. 2021A First-Occupancy Representation for Reinforcement Learningcited by 14
  24. 2020The Variational Bandwidth Bottleneck: Stochastic Evaluation on an Information Budgetcited by 13
  25. 2022Generalised Policy Improvement with Geometric Policy Compositioncited by 13
  26. 2018The Deep Weight Priorcited by 11
  27. 2021Towards an Understanding of Default Policies in Multitask Policy Optimizationcited by 8
  28. 2022Joint Representation Training in Sequential Tasks with Shared Structurecited by 5
  29. 2017Learning Approximately Objective Priorscited by 3

Built on by

Not cited yet.