# Plug and Play, Model-Based Reinforcement Learning

arXiv: https://arxiv.org/abs/2108.08960 · HTML: https://mlsanity.com/paper/2108.08960/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 13 references that are in the map.
- First posted on arXiv 20 Aug 2021.

## Builds on (most cited first, 8 of 8)

- [Interaction Networks for Learning about Objects, Relations and Physics](https://mlsanity.com/paper/1612.00222/) (2016)
- [Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning](https://mlsanity.com/paper/1907.04490/) (2019)
- [PHYRE: A New Benchmark for Physical Reasoning](https://mlsanity.com/paper/1908.05656/) (2019)
- [Schema Networks: Zero-shot Transfer with a Generative Causal Model of Intuitive Physics](https://mlsanity.com/paper/1706.04317/) (2017)
- [COBRA: Data-Efficient Model-Based RL through Unsupervised Object Discovery and Curiosity-Driven Exploration](https://mlsanity.com/paper/1905.09275/) (2019)
- [Towards Deep Symbolic Reinforcement Learning](https://mlsanity.com/paper/1609.05518/) (2016)
- [Blocks and Fuel: Frameworks for deep learning](https://mlsanity.com/paper/1506.00619/) (2015)
- [Model-based Multi-Agent Reinforcement Learning with Cooperative Prioritized Sweeping](https://mlsanity.com/paper/2001.07527/) (2020)

## Abstract (arXiv)

Sample-efficient generalisation of reinforcement learning approaches have always been a challenge, especially, for complex scenes with many components. In this work, we introduce Plug and Play Markov Decision Processes, an object-based representation that allows zero-shot integration of new objects from known object classes. This is achieved by representing the global transition dynamics as a union of local transition functions, each with respect to one active object in the scene. Transition dynamics from an object class can be pre-learnt and thus would be ready to use in a new environment. Each active object is also endowed with its reward function. Since there is no central reward function, addition or removal of objects can be handled efficiently by only updating the reward functions of objects involved. A new transfer learning mechanism is also proposed to adapt reward function in such cases. Experiments show that our representation can achieve sample-efficiency in a variety of set-ups.