Methods › Natural Language Processing › Synthesized Attention Mechanisms › Factorized Dense Synthesized Attention

Factorized Dense Synthesized Attention

1 paper tagged archive 2025-07-28

Introduced by Yi Tay et al. in Synthesizer: Rethinking Self-Attention in Transformer Models

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

Factorized Dense Synthesized Attention is a synthesized attention mechanism, similar to dense synthesized attention, but we factorize the outputs to reduce parameters and prevent overfitting. It was proposed as part of the Synthesizer architecture. The factorized variant of the dense synthesizer can be expressed as follows:

A, B = F_A(Xᵢ), F_B(Xᵢ)

where F_A(.) projects input Xᵢ into a dimensions, F_B(.) projects Xᵢ to b dimensions, and a x b = l. The output of the factorized module is now written as:

Y = Softmax(C)G(X)

where C = H_A(A) * H_B(B), where H_A, H_B are tiling functions and C ∈ℝ^(l x l). The tiling function simply duplicates the vector k times, i.e., ℝˡ →ℝˡᵏ. In this case, H_A() is a projection of ℝᵃ →ℝᵃᵇ and H_B() is a projection of ℝᵇ →ℝᵇᵃ. To avoid having similar values within the same block, we compose the outputs of H_A and H_B.

PaperSource

Papers archive 2025-07-28

1 shown of 1, 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.

Tasks archive 2025-07-28

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

TaskPapers
Abstractive Text Summarization1
Dialogue Generation1
Document Summarization1
Language Modeling1
Language Modelling1
Linguistic Acceptability1
Machine Translation1
Semantic Textual Similarity1
Text Generation1
Translation1

Usage over time archive 2025-07-28

Papers per year tagged with Factorized Dense Synthesized Attention: 2020 to 2020, peak 1 1 0 2020: 1 paper 2020
Papers per year the archive tags with this method, by the paper's archive date (1 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

Synthesized Attention MechanismsAttention Mechanisms

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