Methods › General › Attention Modules › Single-Headed Attention

Single-Headed Attention

6 papers tagged archive 2025-07-28

Introduced by Stephen Merity in Single Headed Attention RNN: Stop Thinking With Your Head

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

Single-Headed Attention is a single-headed attention module used in the SHA-RNN language model. The principle design reasons for single-headedness were simplicity (avoiding running out of memory) and scepticism about the benefits of using multiple heads.

Paper Source See Code · Smerity/sha-rnn

Papers archive 2025-07-28

6 shown of 6, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.

Meta-Learning-Based Delayless Subband Adaptive Filter using Complex Self-Attention for Active Noise Control
27 Dec 2024 · 0 repositories · arXiv:2412.19471
SHAQ: Single Headed Attention with Quasi-Recurrence
18 Aug 2021 · 0 repositories · arXiv:2108.08207
Deep Diacritization: Efficient Hierarchical Recurrence for Improved Arabic Diacritization
1 Nov 2020 · 1 repository · arXiv:2011.00538
Attention-based Joint Detection of Object and Semantic Part
5 Jul 2020 · 1 repository · arXiv:2007.02419

Syntology ran 3 of 12 samples · 9 unverified
Single Headed Attention RNN: Stop Thinking With Your Head
26 Nov 2019 · 5 repositories · arXiv:1911.11423
Attention and Lexicon Regularized LSTM for Aspect-based Sentiment Analysis
1 Jul 2019 · 1 repository

Tasks archive 2025-07-28

17 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

Task	Papers
Sentence	2
Arabic Text Diacritization	1
Aspect-Based Sentiment Analysis	1
Aspect-Based Sentiment Analysis (ABSA)	1
Decoder	1
GPU	1
Hyperparameter Optimization	1
Language Modeling	1
Language Modelling	1
Meta-Learning	1
Object	1
Object Detection	1
Semantic Part Detection	1
Sentiment Analysis	1
object-detection	1
regression	1
valid	1

Usage over time archive 2025-07-28

Papers per year the archive tags with this method, by the paper's archive date (6 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Attention Modules

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections