AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

On the Dynamics of Training Attention Models

Haoye Lu, Yongyi Mao, Amiya Nayak

ICLR 2021 · Architectures and efficiency › Attention and transformers · approach: Theory of self-attention dynamics · 2 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 the full attention model instead of freezing topic-word scores (ablation)attnTC · Synthetic trainingtraining loss decrease speedImproved
Use the full attention model instead of fixing topic-word embeddings (ablation)attnTC · Synthetic trainingtraining loss decrease speedImproved
Use a lower-capacity classifierattention models · Synthetic datasettopic-word score and embedding normImproved
Train with a trainable query rather than a fixed queryattention models · Synthetic datasetpositive score–embedding norm relationshipNo clear effect
Add parallel CNN preprocessing for embeddings and keysattnTC · Synthetic word-pair taskpositive score–embedding norm relationshipNo 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