AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Systematic Generalization with Edge Transformers

Leon Bergen, Timothy J. O'Donnell, Dzmitry Bahdanau

NeurIPS 2021 · Architectures and efficiency › Attention and transformers · approach: Transformer expressivity and learnability · 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
value ablation: use only edge (i,l) for triangle values (ablation)Edge Transformer · CLUTRR train k=2,3, test k=6accuracy: 82.0 -> 54.4Hurt
value ablation (ablation)Edge Transformer · CLUTRR train k=2,3,4, test k=10accuracy: 92.1 -> 75.2Hurt
value ablation (ablation)Edge Transformer · CFQ dependency parsing MCD3EM: 89.1 -> 84.5Hurt
value ablation (ablation)Edge Transformer · COGSgeneralization accuracy: 87.4 -> 87.8No clear effect
attention ablation: key from x_ij instead of x_lj (ablation)Edge Transformer · CLUTRR train k=2,3,4, test k=10accuracy: 92.1 -> 78.7Hurt
attention ablation (ablation)Edge Transformer · COGSgeneralization accuracy: 87.4 -> 86.3Hurt
attention ablation (ablation)Edge Transformer · CFQ dependency parsing MCD3EM: 89.1 -> 88.1No clear effect
attention ablation (ablation)Edge Transformer · CLUTRR train k=2,3, test k=6accuracy: 82.0 -> 84.4Mixed
untied weights across layers (ablation)Edge Transformer · CFQ dependency parsing MCD3EM: 89.1 -> 37.2Hurt
untied weights across layers (ablation)Edge Transformer · CLUTRR train k=2,3, test k=10accuracy: 58.1 -> 55.3Hurt
untied weights across layers (ablation)Edge Transformer · CLUTRR train k=2,3,4, test k=6accuracy: 98.1 -> 98.2No clear effect
untied weights across layers (ablation)Edge Transformer · COGSgeneralization accuracy: 87.4 -> 86.3No 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