Learning of communication codes in multi-agent reinforcement learning problem

By T. Kasai, Hiroshi Tenmoto, Akimoto Kamiya. 2008

35 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Natural Language Processing, Reinforcement Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Sergey Levine
2. Ali Farhadi
3. Mohit Bansal
4. Shimon Whiteson
5. Aniruddha Kembhavi
6. Yoshua Bengio
7. Pieter Abbeel
8. Ruslan Salakhutdinov

In Natural Language Processing

1. Jason Aaron Edward Weston
2. Mohit Bansal
3. Ali Farhadi
4. Jakob N. Foerster
5. Arthur Szlam
6. Aniruddha Kembhavi
7. Kyunghyun Cho
8. Rob Fergus

In Reinforcement Learning

1. Shimon Whiteson
2. Jakob N. Foerster
3. Rob Fergus
4. Arthur Szlam
5. Sainbayar Sukhbaatar
6. Klaus-Jürgen Diepold
7. Sven Gronauer
8. Davood Hajinezhad

Citations in the corpus

• Learning to Communicate with Deep Multi-Agent Reinforcement Learning (2016, background): “[7], in which tabular Q-learning agents have to learn the content of a message to solve a predator-prey task with communication.”
• Learning to Play Guess Who? and Inventing a Grounded Language as a Consequence (2016, background): “In [9] the predator-prey environment is studied where agents learn a communication protocol consisting of bits.”
• Learning Multiagent Communication with Backpropagation (2016, methodology): “Both tasks are trained for 300 epochs, each epoch being 100 weight updates with RMSProp [32] on mini-batch of 288 game episodes (distributed over multiple CPU cores).”
• Learning to Communicate to Solve Riddles with Deep Distributed Recurrent Q-Networks (2016, background): “One exception is the work of Kasai et al. (2008), in which the tabular Q-learning agents have to learn the content of a message to solve a predator-prey task.”
• Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithms (2019, background): “For example, [128, 21, 129] investigate learning to communicate, in order for the agents to better coordinate.”
• Multi-agent deep reinforcement learning: a survey (2021, background): “Early MARL works investigated communication rooted in tabular worlds with limited observability (Kasai et al. 2008).”
• A review of cooperative multi-agent deep reinforcement learning (2019, background): “[61] proposed one of first learning to communicate algorithms, in which tabular Q-learning agents learn messages to communicate with other agents in the predator-prey environment.”
• A Cordial Sync: Going Beyond Marginal Policies for Multi-Agent Embodied Tasks (2020, background): “Similarly, communication and language in the multi-agent setting has been investigated [32,44,10,60,52,27,79,66,7] in maze-based setups, tabular tasks, and Markov games.”