← Back to timeline
Deep Learning1997demo planned

LSTM

Sequence memory with gates

LSTM made recurrent networks much better at carrying information across long sequences by controlling what to keep, write, and forget.

Roadmap node · demo on the way

Read the wiki entry for LSTM

The interactive demo for this model is on the build list — the live linear regression demo is the template the rest will follow. The wiki entry already has the full concept, history, and where this fits in the chain.