Universally Expressive Communication in Multi-Agent Reinforcement Learning

Oct-9-2025, 16:41:05 GMT–Neural Information Processing Systems

By observing that many existing protocols can be viewed as instances of graph neural networks (GNNs), we demonstrate the equivalence of joint action selection to node labelling.

communication, machine learning, reinforcement learning, (15 more...)

Neural Information Processing Systems

Oct-9-2025, 16:41:05 GMT

Conferences PDF

Country:
- Europe > United Kingdom > England
  - Cambridgeshire > Cambridge (0.04)
  - Oxfordshire > Oxford (0.04)

Genre:
- Research Report (0.69)

Technology:
- Information Technology > Artificial Intelligence
  - Machine Learning
    - Learning Graphical Models > Undirected Networks
      - Markov Models (0.46)
    - Neural Networks > Deep Learning (0.46)
    - Reinforcement Learning (1.00)
  - Representation & Reasoning > Agents (1.00)