AI Paper Atlas

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

Emergent Communication for Rules Reasoning

Yuxuan Guo, Yifan Hao, Rui Zhang, Enshuai Zhou, Zidong Du, Xishan Zhang, Xinkai Song, Yuanbo Wen, Yongwei Zhao, Xuehai Zhou, Jiaming Guo, Qi Yi and 5 more

NeurIPS 2023 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Emergent communication in referential games · 0 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
add two-stage curriculum training (speaker pre-training on held-out rules) (ablation)Reasoning Game agents · rule-RAVEN N=20/30/40train accuracy: ~0.9 vs low accuracy similar to message-blockedImproved
generate candidates with rule-based distract rules instead of I-RAVEN style (ablation)rule-RAVEN vs I-RAVEN · message-blocked listenerlistener train accuracy (overfitting): ~0.9 -> ~0.6Improved
increase attribute values from 20 to 30 or 40 in I-RAVEN (ablation)I-RAVEN · message-blocked listeneroverfittingNo clear effect
measure Rule-Message vs Panel-Message Topsimemerged language · ID split N=20Topsim: 0.1148 -> 0.2382Improved
increase attribute value diversity N from 20 to 40Reasoning Game agents · ID accuracyaccuracy: 0.9492 -> 0.9194No clear effect
generalize to unseen rule combinations (Inpo-ood vs ID)emerged language · N=20accuracy: 0.9492 -> 0.9473No clear effect
rules seen on other attributes (Expo-ood-L1 vs Expo-ood-L2)emerged language · N=20accuracy: 0.4921 -> 0.6590Improved
harder generalization from Inpo-ood to Expo-ood-L1 to Expo-ood-L2emerged language · averaged over Naccuracy: 0.92 -> 0.58 -> 0.46Hurt
transfer emerged language from source N=20 to target N=80new listeners · task transfertrain accuracy: 0.9199No clear effect
emerged language vs random mapping for transfernew listeners · task transfer curvesconvergenceImproved
emerged language vs ideal rule-encoding language for transfernew listeners · task transfer curvesconvergence speedNo clear effect
longer messages (M=6) and larger vocabulary (V=30) (ablation)Reasoning Game agents · ID accuracy N=20accuracy: 0.9539 -> 0.9522No clear effect

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