AI Paper Atlas

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

Emergent Communication: Generalization and Overfitting in Lewis Games

Mathieu Rita, Corentin Tallec, Paul Michel, Jean-Bastien Grill, Olivier Pietquin, Emmanuel Dupoux, Florian Strub

NeurIPS 2022 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Emergent communication in referential games · 3 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
early-stopped listener instead of continuous listener (ablation)Lewis game agents · Reconstruction Lewis gameGeneralization: 0.58 -> 0.95Improved
early-stopped listener instead of continuous listener (ablation)Lewis game agents · Reconstruction Lewis gameCompositionality: 0.22 -> 0.39Improved
top partial listener (tuned N_step) instead of continuous listener (ablation)Lewis game agents · Reconstruction Lewis gameTest co-adaptation loss: 4.64 -> 0.97Improved
add dropout to listener (ablation)Continuous listener · Reconstruction Lewis gameGeneralization: 0.58 -> 0.64Improved
remove layer norm from listener (ablation)Continuous listener · Reconstruction Lewis gameGeneralization: 0.58 -> 0.70Improved
add weight decay to listener (ablation)Continuous listener · Reconstruction Lewis gameGeneralization: 0.58 -> 0.72Improved
remove layer norm and add weight decay (ablation)Continuous listener · Reconstruction Lewis gameGeneralization: 0.58 -> 0.87Improved
dropout on listener (ablation)Continuous listener · Reconstruction Lewis gameTest co-adaptation loss: 4.64 -> 4.86No clear effect
early-stopped listener on CelebA 1/100 subset (ablation)Discrimination game · CelebA 1/100Generalization: 0.39 -> 0.69Improved
early-stopped listener on CelebA full data (ablation)Discrimination game · CelebA 1/1Generalization: 0.94 -> 0.97Improved
early-stopped listener on ImageNet 1/100 subset (ablation)Discrimination game · ImageNet 1/100Generalization: 0.51 -> 0.64Improved
early-stopped listener on ImageNet full data (ablation)Discrimination game · ImageNet 1/1Generalization: 0.96 -> 0.98Improved

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