Gu et al: Efficiently Modeling Long Sequences with Structured State Spaces [80ec42fa]
- papers: arxiv.org/pdf/2111.00396.pdf
- Video: Efficiently Modeling Long Sequences with Structured State Spaces
-
Open review version: openreview.net/pdf?id=uYLFoz1vlAC
- review notes: openreview.net/forum?id=uYLFoz1vlAC
- state space continious representation: en.wikipedia.org/wiki/State-space_representation
- previous work: openreview.net/pdf?id=yWd42CWN3c
-
based off of hippo: proceedings.neurips.cc/paper/2020/hash/102f0bb6efb3a6128a3c750dd16729be-Abstract.html
- papertalk.org/papertalks/9174
- github.com/HazyResearch/hippo-code
- proceedings.neurips.cc/paper/2020/file/102f0bb6efb3a6128a3c750dd16729be-Paper.pdf
- crossminds.ai/video/hippo-recurrent-memory-with-optimal-polynomial-projections-606fddf7f43a7f2f827bf91d/
- proceedings.neurips.cc/paper/2020/file/102f0bb6efb3a6128a3c750dd16729be-Paper.pdf
1. Past Similar Work: Legendre Memory Unit
1. Past Similar Work: Legendre Memory Unit
- www.youtube.com/watch?v=8t64QaTdBcU
- proceedings.neurips.cc/paper/2019/file/952285b9b7e7a1be5aa7849f32ffff05-Paper.pdf
- arxiv.org/abs/2102.11417
- en.wikipedia.org/wiki/Legendre_polynomials#Definition_by_construction_as_an_orthogonal_system