AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Emergent Discrete Communication in Semantic Spaces

Mycal Tucker, Huao Li, Siddharth Agrawal, Dana Hughes, Katia P. Sycara, Michael Lewis, Julie Shah

NeurIPS 2021 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Emergent communication in referential games · 7 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use learnable prototype tokens instead of one-hot tokensMulti-agent communication · Triangle, uniform prior, σ=0.9, c=9self-play reward: -201 -> -124Improved
Use learnable prototype tokens instead of one-hot tokensMulti-agent communication · Triangle, skewed prior, σ=0.9, c=9self-play reward: -136 -> -100Improved
Use learnable prototype tokens in a 3-dimensional channel for 9 targets9 Points · 9 Points, σ=0, c=3self-play reward: -96 -> -60Improved
Use learnable prototype tokens instead of one-hot tokens9 Points · 9 Points, σ=0.5, c=9self-play reward: -132 -> -100Improved
Use prototype tokens instead of continuous communicationUniform · Uniformreward: -97.8 -> -113.6Hurt
Use prototype tokens instead of one-hot tokensCIFAR10 reference game · Label Outaccuracy: 50% (0.8) -> 62% (1.2)Improved
Use prototype communication with BERT embeddings instead of continuous communicationCIFAR10 reference game · AMT Outaccuracy: 53% (0.6) -> 56% (0.6)Improved
Use prototype tokens instead of BERT-based embeddings as the human-interpreted communicationHuman listeners · Human understanding of held-out tokensaccuracy: 66% -> 70%No clear effect
Remove noise from prototype-agent training (ablation)CIFAR10 reference game · Label Outaccuracy: 58% (1.0) -> 55% (1.0)Hurt
Remove non-identity reward structure from prototype-agent training (ablation)CIFAR10 reference game · Label Outaccuracy: 58% (1.0) -> 53% (1.0)Hurt
Remove both communication noise and non-identity reward structure (ablation)CIFAR10 reference game · AMT Outaccuracy: 56% (0.6) -> 53% (0.5)Hurt
Replace class-name embeddings with animal/vehicle binary labels (ablation)Proto (BERT) · CIFAR10 reference gameLabel In accuracy: 93% (0.7) -> 61% (0.3)Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy