mlsanityData to 7 Oct 2026

Natural Language Does Not Emerge 'Naturally' in Multi-Agent Dialog

arXiv 1706.08502 ↗First posted 26 Jun 2017cs.CL cs.AI cs.CVcited by 86

Not on the map. Why?

Abstract (arXiv)

A number of recent works have proposed techniques for end-to-end learning of communication protocols among cooperative multi-agent populations, and have simultaneously found the emergence of grounded human-interpretable language in the protocols developed by the agents, all learned without any human supervision! In this paper, using a Task and Tell reference game between two agents as a testbed, we present a sequence of 'negative' results culminating in a 'positive' one -- showing that while most agent-invented languages are effective (i.e. achieve near-perfect task rewards), they are decidedly not interpretable or compositional. In essence, we find that natural language does not emerge 'naturally', despite the semblance of ease of natural-language-emergence that one may gather from recent literature. We discuss how it is possible to coax the invented languages to become more and more human-like and compositional by increasing restrictions on how two agents may communicate.

Key facts. Cited by 86 papers in the map (data through 7 Oct 2026). First posted on arXiv 26 Jun 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 21; 2022: 16; 2023: 16; 2024: 14; 2025: 8; 2026: 11.

86 papers.

Sort:

By area:

  1. 2022What Artificial Neural Networks Can Tell Us About Human Language Acquisitioncited by 97
  2. 2021Meta-Learning to Compositionally Generalizecited by 62
  3. 2024Learning from Teaching Regularization: Generalizable Correlations Should be Easy to Imitatecited by 46
  4. 2021Deep Multiagent Reinforcement Learning: Challenges and Directionscited by 42
  5. 2021Emergent Communication of Generalizationscited by 41
  6. 2022Emergent Communication: Generalization and Overfitting in Lewis Gamescited by 28
  7. 2021Interactive Learning from Activity Descriptioncited by 26
  8. 2021Interpretable agent communication from scratch (with a generic visual processor emerging on the side)cited by 25
  9. 2022Emergent Communication for Understanding Human Language Evolution: What's Missing?cited by 23
  10. 2021Iterated learning for emergent systematicity in VQAcited by 21
  11. 2021Emergent Discrete Communication in Semantic Spacescited by 20
  12. 2022Neuro-Symbolic Causal Reasoning Meets Signaling Game for Emergent Semantic Communicationscited by 20
38 more
  1. 2021Quasi-Equivalence Discovery for Zero-Shot Emergent Communicationcited by 19
  2. 2021Interpretation of Emergent Communication in Heterogeneous Collaborative Embodied Agentscited by 19
  3. 2024Iterated Learning Improves Compositionality in Large Vision-Language Modelscited by 17
  4. 2024A Review of the Applications of Deep Learning-Based Emergent Communicationcited by 17
  5. 2024Language Grounded Multi-agent Reinforcement Learning with Human-interpretable Communicationcited by 17
  6. 2024Generative Emergent Communication: Large Language Model is a Collective World Modelcited by 15
  7. 2022Learning to Mediate Disparities Towards Pragmatic Communicationcited by 14
  8. 2024Efficacy of Language Model Self-Play in Non-Zero-Sum Gamescited by 14
  9. 2021The Emergence of the Shape Bias Results from Communicative Efficiencycited by 13
  10. 2023What makes a language easy to deep-learn? Deep neural networks and humans similarly benefit from compositional structurecited by 13
  11. 2021Catalytic Role Of Noise And Necessity Of Inductive Biases In The Emergence Of Compositional Communicationcited by 12
  12. 2023Towards More Human-like AI Communication: A Review of Emergent Communication Researchcited by 12
  13. 2023Lewis's Signaling Game as beta-VAE For Natural Word Lengths and Segmentscited by 11
  14. 2025The Collaboration Gap: Exploration and Benchmarking of Open-World Agentic Cooperationcited by 11
  15. 2021Dynamic population-based meta-learning for multi-agent communication with natural languagecited by 10
  16. 2022Reinforcement Learning and Bandits for Speech and Language Processing: Tutorial, Review and Outlookcited by 10
  17. 2021Multitasking Inhibits Semantic Driftcited by 8
  18. 2022Emergence of hierarchical reference systems in multi-agent communicationcited by 8
  19. 2022Defending Compositionality in Emergent Languagescited by 8
  20. 2024NeLLCom-X: A Comprehensive Neural-Agent Framework to Simulate Language Learning and Group Communicationcited by 8
  21. 2022FCMNet: Full Communication Memory Net for Team-Level Cooperation in Multi-Agent Systemscited by 7
  22. 2024The Curious Case of Representational Alignment: Unravelling Visio-Linguistic Tasks in Emergent Communicationcited by 7
  23. 2021The Effect of Efficient Messaging and Input Variability on Neural-Agent Iterated Language Learningcited by 6
  24. 2021The emergence of visual semantics through communication gamescited by 5
  25. 2023Visual Referential Games Further the Emergence of Disentangled Representationscited by 5
  26. 2023Metropolis-Hastings algorithm in joint-attention naming game: Experimental semiotics studycited by 5
  27. 2024Language Evolution with Deep Learningcited by 5
  28. 2021Neural networks can understand compositional functions that humans do not, in the context of emergent communicationcited by 4
  29. 2022Over-communicate no more: Situated RL agents learn concise communication protocolscited by 4
  30. 2023Learning Multi-Object Positional Relationships via Emergent Communicationcited by 4
  31. 2021TexRel: a Green Family of Datasets for Emergent Communications on Relationscited by 3
  32. 2022Emergence of Machine Language: Towards Symbolic Intelligence with Neural Networkscited by 3
  33. 2023Do learned speech symbols follow Zipf's law?cited by 3
  34. 2023ChapGTP, ILLC's Attempt at Raising a BabyLM: Improving Data Efficiency by Automatic Task Formationcited by 3
  35. 2025Unsupervised Translation of Emergent Communicationcited by 3
  36. 2021Zero-Shot Generalization using Intrinsically Motivated Compositional Emergent Protocolscited by 2
  37. 2021DORA: Toward Policy Optimization for Task-oriented Dialogue System with Efficient Contextcited by 2
  38. 2022Recommendations for Systematic Research on Emergent Languagecited by 2

50 most cited of 86 All 86 (JSON)