Miss Tools and Mr Fruit: Emergent Communication in Agents Learning about Object Affordances

Recent research studies communication emergence in communities of deep network agents assigned a joint task, hoping to gain insights on human language evolution. We propose here a new task capturing crucial aspects of the human environment, such as natural object affordances, and of human conversation, such as full symmetry among the participants. By conducting a thorough pragmatic and semantic analysis of the emergent protocol, we show that the agents solve the shared task through genuine bilateral, referential communication. However, the agents develop multiple idiolects, which makes us conclude that full symmetry is not a sufficient condition for a common language to emerge.

[1]  Yann Dauphin,et al.  Deal or No Deal? End-to-End Learning of Negotiation Dialogues , 2017, EMNLP.

[2]  Ronald J. Williams,et al.  Simple Statistical Gradient-Following Algorithms for Connectionist Reinforcement Learning , 2004, Machine Learning.

[3]  Simon Kirby,et al.  Situated Cognition and the Role of Multi-agent Models in Explaining Language Structure , 2002, Adaptive Agents and Multi-Agents Systems.

[4]  Stephen Clark,et al.  Emergent Communication through Negotiation , 2018, ICLR.

[5]  Yi Wu,et al.  Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments , 2017, NIPS.

[6]  Marco Baroni,et al.  How agents see things: On visual representations in an emergent language game , 2018, EMNLP.

[7]  Moisès Esteban-Guitart.  A natural history of human thinking Michael Tomasello A natural history of human thinking , 2014, Animal Behaviour.

[8]  Carina Silberer,et al.  Models of Semantic Representation with Visual Attributes , 2013, ACL.

[9]  Mark S. Seidenberg,et al.  Semantic feature production norms for a large set of living and nonliving things , 2005, Behavior research methods.

[10]  L. Steels.  Evolving grounded communication for robots , 2003, Trends in Cognitive Sciences.

[11]  Wentian Li,et al.  Zipf's Law everywhere , 2002, Glottometrics.

[12]  J. Pearl,et al.  Causal Inference in Statistics: A Primer , 2016 .

[13]  Kyunghyun Cho,et al.  Emergent Communication in a Multi-Modal, Multi-Step Referential Game , 2017, ICLR.

[14]  Nando de Freitas,et al.  Intrinsic Social Motivation via Causal Influence in Multi-Agent RL , 2018, ArXiv.

[15]  Yu-Jung Heo,et al.  Answerer in Questioner's Mind for Goal-Oriented Visual Dialogue , 2018, ArXiv.

[16]  Laura Graesser,et al.  Emergent Linguistic Phenomena in Multi-Agent Communication Games , 2019, EMNLP.

[17]  Joaquin Quiñonero Candela,et al.  Counterfactual reasoning and learning systems: the example of computational advertising , 2013, J. Mach. Learn. Res..

[18]  Emil Gustavsson,et al.  Learning to Play Guess Who? and Inventing a Grounded Language as a Consequence , 2016, ArXiv.

[19]  José M. F. Moura,et al.  Natural Language Does Not Emerge ‘Naturally’ in Multi-Agent Dialog , 2017, EMNLP.

[20]  Nando de Freitas,et al.  Compositional Obverter Communication Learning From Raw Visual Input , 2018, ICLR.

[21]  Territoire Urbain,et al.  Convention , 1955, Hidden Nature.

[22]  Alexander Peysakhovich,et al.  Multi-Agent Cooperation and the Emergence of (Natural) Language , 2016, ICLR.

[23]  Stefan Lee,et al.  Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[24]  David Reitter,et al.  How groups develop a specialized domain vocabulary: A cognitive multi-agent model , 2011, Cognitive Systems Research.

[25]  Stephen Clark,et al.  Emergence of Linguistic Communication from Referential Games with Symbolic and Pixel Input , 2018, ICLR.

[26]  Ivan Titov,et al.  Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols , 2017, NIPS.

[27]  Pieter Abbeel,et al.  Emergence of Grounded Compositional Language in Multi-Agent Populations , 2017, AAAI.