Incorporating Pragmatic Reasoning Communication into Emergent Language
arXiv:2006.04109
Abstract
Emergentism and pragmatics are two research fields that study the dynamics of linguistic communication along substantially different timescales and intelligence levels. From the perspective of multi-agent reinforcement learning, they correspond to stochastic games with reinforcement training and stage games with opponent awareness. Given that their combination has been explored in linguistics, we propose computational models that combine short-term mutual reasoning-based pragmatics with long-term language emergentism. We explore this for agent communication referential games as well as in Starcraft II, assessing the relative merits of different kinds of mutual reasoning pragmatics models both empirically and theoretically. Our results shed light on their importance for making inroads towards getting more natural, accurate, robust, fine-grained, and succinct utterances.
arXiv comment9 pages. Accepted as a spotlight paper to NeurIPS 2020
References in corpus (14)
- Learning to Communicate with Deep Multi-Agent Reinforcement Learning
- Counterfactual Multi-Agent Policy Gradients
- Autonomous Agents Modelling Other Agents: A Comprehensive Survey and Open Problems
- QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning
- A Survey of Learning in Multiagent Environments: Dealing with Non-Stationarity
- Multi-Agent Cooperation and the Emergence of (Natural) Language
- Learning to Communicate to Solve Riddles with Deep Distributed Recurrent Q-Networks
- Bayesian Action Decoder for Deep Multi-Agent Reinforcement Learning
- A Rate-Distortion view of human pragmatic reasoning
- Compositional Obverter Communication Learning From Raw Visual Input
- Emergent Translation in Multi-Agent Communication
- A mathematical theory of cooperative communication
- Pragmatically Informative Text Generation
- Learning to Communicate in Multi-Agent Reinforcement Learning : A Review