Skip to yearly menu bar Skip to main content


Poster

StableSSM: Alleviating the Curse of Memory in State-space Models through Stable Reparameterization

Shida Wang · Qianxiao Li

Hall C 4-9 #1006
[ ] [ Paper PDF ]
Tue 23 Jul 4:30 a.m. PDT — 6 a.m. PDT

Abstract:

In this paper, we investigate the long-term memory learning capabilities of state-space models (SSMs) from the perspective of parameterization. We prove that state-space models without any reparameterization exhibit a memory limitation similar to that of traditional RNNs: the target relationships that can be stably approximated by state-space models must have an exponential decaying memory. Our analysis identifies this ``curse of memory'' as a result of the recurrent weights converging to a stability boundary, suggesting that a reparameterization technique can be effective. To this end, we introduce a class of reparameterization techniques for SSMs that effectively lift its memory limitations. Besides improving approximation capabilities, we further illustrate that a principled choice of reparameterization scheme can also enhance optimization stability. We validate our findings using synthetic datasets, language models and image classifications.

Chat is not available.