mlsanityData to 7 Oct 2026

Reinforcement Learning through Active Inference

arXiv 2002.12636 ↗First posted 28 Feb 2020cs.LG cs.AI cs.ITcited by 47

Not on the map. Why?

Abstract (arXiv)

The central tenet of reinforcement learning (RL) is that agents seek to maximize the sum of cumulative rewards. In contrast, active inference, an emerging framework within cognitive and computational neuroscience, proposes that agents act to maximize the evidence for a biased generative model. Here, we illustrate how ideas from active inference can augment traditional RL approaches by (i) furnishing an inherent balance of exploration and exploitation, and (ii) providing a more flexible conceptualization of reward. Inspired by active inference, we develop and implement a novel objective for decision making, which we term the free energy of the expected future. We demonstrate that the resulting algorithm successfully balances exploration and exploitation, simultaneously achieving robust performance on several challenging RL benchmarks with sparse, well-shaped, and no rewards.

Key facts. Cited by 47 papers in the map (data through 7 Oct 2026). First posted on arXiv 28 Feb 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 15; 2022: 7; 2023: 5; 2024: 6; 2025: 7; 2026: 7.

47 papers.

Sort:

By area:

  1. 2021Predictive Coding: a Theoretical and Experimental Reviewcited by 74
  2. 2021Active Inference in Robotics and Artificial Agents: Survey and Challengescited by 52
  3. 2022pymdp: A Python library for active inference in discrete state spacescited by 39
  4. 2024At the Dawn of Generative AI Era: A Tutorial-cum-Survey on New Frontiers in 6G Wireless Intelligencecited by 25
  5. 2022The Free Energy Principle for Perception and Action: A Deep Learning Perspectivecited by 24
  6. 2021Contrastive Active Inferencecited by 21
  7. 2021Backprop-Free Reinforcement Learning with Active Neural Generative Codingcited by 12
  8. 2021Understanding the Origin of Information-Seeking Exploration in Probabilistic Objectives for Controlcited by 10
  9. 2024R-AIF: Solving Sparse-Reward Robotic Tasks from Pixels with Active Inference and World Modelscited by 10
  10. 2021Applications of the Free Energy Principle to Machine Learning and Neurosciencecited by 9
  11. 2022Active Exploration for Robotic Manipulationcited by 9
  12. 2021Prior Preference Learning from Experts:Designing a Reward with Active Inferencecited by 8
35 more
  1. 2022Exploration via Planning for Information about the Optimal Trajectorycited by 8
  2. 2022Complex behavior from intrinsic motivation to occupy action-state path spacecited by 7
  3. 2021Explore the Context: Optimal Data Collection for Context-Conditional Dynamics Modelscited by 6
  4. 2021Active Inference and Epistemic Value in Graphical Modelscited by 6
  5. 2024Adaptive Stream Processing on Edge Devices through Active Inferencecited by 6
  6. 2021Chance-Constrained Active Inferencecited by 5
  7. 2021Online reinforcement learning with sparse rewards through an active inference capsulecited by 3
  8. 2022Successor Representation Active Inferencecited by 3
  9. 2023Active Inference in String Diagrams: A Categorical Account of Predictive Processing and Free Energycited by 3
  10. 2024Intrinsic Rewards for Exploration without Harm from Observational Noise: A Simulation Study Based on the Free Energy Principlecited by 3
  11. 2021Active Inference Tree Search in Large POMDPscited by 2
  12. 2021Deep Active Inference for Pixel-Based Discrete Control: Evaluation on the Car Racing Problemcited by 2
  13. 2022Active Inference for Robotic Manipulationcited by 2
  14. 2023Sophisticated Learning: A novel algorithm for active learning during model-based planningcited by 2
  15. 2025The Missing Reward: Active Inference in the Era of Experiencecited by 2
  16. 2021A Mathematical Walkthrough and Discussion of the Free Energy Principlecited by 1
  17. 2023Life-inspired Interoceptive Artificial Intelligence for Autonomous and Adaptive Agentscited by 1
  18. 2024Free Energy Projective Simulation (FEPS): Active inference with interpretabilitycited by 1
  19. 2025A Framework for Inherently Safer AGI through Language-Mediated Active Inferencecited by 1
  20. 2021Making AI 'Smart': Bridging AI and Cognitive Sciencecited by 0
  21. 2023A Neural Network Implementation for Free Energy Principlecited by 0
  22. 2023Tactile Active Inference Reinforcement Learning for Efficient Robotic Manipulation Skill Acquisitioncited by 0
  23. 2024Non-maximizing policies that fulfill multi-criterion aspirations in expectationcited by 0
  24. 2025Boosting MCTS with Free Energy Minimizationcited by 0
  25. 2025Active Multimodal Distillation for Few-shot Action Recognitioncited by 0
  26. 2025UF-RNN: Real-Time Adaptive Motion Generation Using Uncertainty-Driven Foresight Predictioncited by 0
  27. 2025Prior preferences in active inference agents: soft, hard, and goal shapingcited by 0
  28. 2025EcoNet: Multiagent Planning and Control Of Household Energy Resources Using Active Inferencecited by 0
  29. 2026Distributional Active Inferencecited by 0
  30. 2026BRAIN: Bayesian Reasoning via Active Inference for Agentic and Embodied Intelligence in Mobile Networkscited by 0
  31. 2026Active Inference: A method for Phenotyping Agency in AI systems?cited by 0
  32. 2026Mind Dreamer: Untethering Imagination via Active Causal Intervention on Latent Manifoldscited by 0
  33. 2026Exploratory Experience Shapes the Geometry of Predictive Representationscited by 0
  34. 2026Expected Free Energy as Belief-Dependent Utility for rho-POMDPscited by 0
  35. 2026Expected Free Energy-based Informative Path Planning for Robotic Mars Explorationcited by 0