mlsanityData to 14 Oct 2026

Learning to Perform Physics Experiments via Deep Reinforcement Learning

arXiv 1611.01843 ↗First posted 6 Nov 2016stat.ML cs.AI cs.CVcited by 47

Not on the map. Why?

Abstract (arXiv)

When encountering novel objects, humans are able to infer a wide range of physical properties such as mass, friction and deformability by interacting with them in a goal driven way. This process of active interaction is in the same spirit as a scientist performing experiments to discover hidden facts. Recent advances in artificial intelligence have yielded machines that can achieve superhuman performance in Go, Atari, natural language processing, and complex control problems; however, it is not clear that these systems can rival the scientific intuition of even a young child. In this work we introduce a basic set of tasks that require agents to estimate properties such as mass and cohesion of objects in an interactive simulated environment where they can manipulate the objects and observe the consequences. We found that state of art deep reinforcement learning methods can learn to perform the experiments necessary to discover such hidden properties. By systematically manipulating the problem difficulty and the cost incurred by the agent for performing experiments, we found that agents learn different strategies that balance the cost of gathering information against the cost of making mistakes in different situations.

Key facts. Cited by 47 papers in the map (data through 14 Oct 2026). First posted on arXiv 6 Nov 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 12; 2022: 8; 2023: 11; 2024: 8; 2025: 6; 2026: 2.

47 papers.

Sort:

By area:

  1. 2024REBEL: Reinforcement Learning via Regressing Relative Rewardscited by 69
  2. 2025A Survey of Sim-to-Real Methods in RL: Progress, Prospects and Challenges with Foundation Modelscited by 38
  3. 2023Plan, Eliminate, and Track -- Language Models are Good Teachers for Embodied Agentscited by 32
  4. 2021Sample Efficient Reinforcement Learning via Model-Ensemble Exploration and Exploitationcited by 18
  5. 2024A Review of Reward Functions for Reinforcement Learning in the context of Autonomous Drivingcited by 18
  6. 2021Sample-efficient Reinforcement Learning Representation Learning with Curiosity Contrastive Forward Dynamics Modelcited by 16
  7. 2021CoMPS: Continual Meta Policy Searchcited by 15
  8. 2021Comparing Popular Simulation Environments in the Scope of Robotics and Reinforcement Learningcited by 12
  9. 2023Gradient Shaping for Multi-Constraint Safe Reinforcement Learningcited by 10
  10. 2024PrimeGuard: Safe and Helpful LLMs through Tuning-Free Routingcited by 10
  11. 2022Hyperparameter Tuning for Deep Reinforcement Learning Applicationscited by 7
  12. 2021Network Defense is Not a Gamecited by 6
35 more
  1. 2022Explainability in Deep Reinforcement Learning, a Review into Current Methods and Applicationscited by 6
  2. 2024Model-Free Active Exploration in Reinforcement Learningcited by 5
  3. 2021Combining machine learning with physics: A framework for tracking and sorting multiple dark solitonscited by 4
  4. 2022Innovations in Integrating Machine Learning and Agent-Based Modeling of Biomedical Systemscited by 4
  5. 2023On the Challenges of using Reinforcement Learning in Precision Drug Dosing: Delay and Prolongedness of Action Effectscited by 4
  6. 2024Stealing That Free Lunch: Exposing the Limits of Dyna-Style Reinforcement Learningcited by 4
  7. 2022A model-based approach to meta-Reinforcement Learning: Transformers and tree searchcited by 3
  8. 2023A Deep Reinforcement Learning Based Motion Cueing Algorithm for Vehicle Driving Simulationcited by 3
  9. 2024Artificial intelligence, rationalization, and the limits of control in the public sector: the case of tax policy optimizationcited by 3
  10. 2021Shifting Capsule Networks from the Cloud to the Deep Edgecited by 2
  11. 2022Reinforcement Learning-based Defect Mitigation for Quality Assurance of Additive Manufacturingcited by 2
  12. 2022Actively Learning Costly Reward Functions for Reinforcement Learningcited by 2
  13. 2023Multi-Agent Reinforcement Learning with Action Masking for UAV-enabled Mobile Communicationscited by 2
  14. 2023A Review of Symbolic, Subsymbolic and Hybrid Methods for Sequential Decision Makingcited by 2
  15. 2024Progressive-Resolution Policy Distillation: Leveraging Coarse-Resolution Simulations for Time-Efficient Fine-Resolution Policy Learningcited by 2
  16. 2025Portable Reward Tuning: Towards Reusable Fine-Tuning across Different Pretrained Modelscited by 2
  17. 2025Scan, Materialize, Simulate: A Generalizable Framework for Physically Grounded Robot Planningcited by 2
  18. 2021Parameter-free Gradient Temporal Difference Learningcited by 1
  19. 2021Secure Bayesian Federated Analytics for Privacy-Preserving Trend Detectioncited by 1
  20. 2022AACHER: Assorted Actor-Critic Deep Reinforcement Learning with Hindsight Experience Replaycited by 1
  21. 2023Learning active tactile perception through belief-space controlcited by 1
  22. 2023Partial End-to-end Reinforcement Learning for Robustness Against Modelling Error in Autonomous Racingcited by 1
  23. 2021Bridge the Vision Gap from Field to Command: A Deep Learning Network Enhancing Illumination and Detailscited by 0
  24. 2021Intuitive Physics Guided Exploration for Sample Efficient Sim2real Transfercited by 0
  25. 2021Off-policy Reinforcement Learning with Optimistic Exploration and Distribution Correctioncited by 0
  26. 2022Towards Robust On-Ramp Merging via Augmented Multimodal Reinforcement Learningcited by 0
  27. 2023Multi-Agent Interplay in a Competitive Survival Environmentcited by 0
  28. 2023Fundamental Limits of Deep Learning-Based Binary Classifiers Trained with Hinge Losscited by 0
  29. 2023Closed Drafting as a Case Study for First-Principle Interpretability, Memory, and Generalizability in Deep Reinforcement Learningcited by 0
  30. 2024A proximal policy optimization based intelligent home solar managementcited by 0
  31. 2025Proposing Hierarchical Goal-Conditioned Policy Planning in Multi-Goal Reinforcement Learningcited by 0
  32. 2025Extending a Quantum Reinforcement Learning Exploration Policy with Flags to Connect Fourcited by 0
  33. 2025Scalable Submodular Policy Optimization via Pruned Submodularity Graphcited by 0
  34. 2026Learning What Matters: Adaptive Information-Theoretic Objectives for Robot Explorationcited by 0
  35. 2026New Evidence, Same Choice: Testing Physical Experiment Selection in Vision Language Modelscited by 0