Pith. sign in

REVIEW

Sketching as a Tool for Understanding and Accelerating Self-attention for Long Sequences

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2112.05359 v1 pith:Q5GIJB5Z submitted 2021-12-10 cs.LG cs.CLstat.ML

classification cs.LGcs.CLstat.ML
keywords self-attentionlongcomplexitymatrixmodelsquadraticsamplingsequences
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Transformer-based models are not efficient in processing long sequences due to the quadratic space and time complexity of the self-attention modules. To address this limitation, Linformer and Informer are proposed to reduce the quadratic complexity to linear (modulo logarithmic factors) via low-dimensional projection and row selection respectively. These two models are intrinsically connected, and to understand their connection, we introduce a theoretical framework of matrix sketching. Based on the theoretical analysis, we propose Skeinformer to accelerate self-attention and further improve the accuracy of matrix approximation to self-attention with three carefully designed components: column sampling, adaptive row normalization and pilot sampling reutilization. Experiments on the Long Range Arena (LRA) benchmark demonstrate that our methods outperform alternatives with a consistently smaller time/space footprint.

Discussion (0). Continue with ORCID to comment.

Pith tools