LSTM

← Back to Recurrent Neural Networks

Long Short-Term Memory. Addresses the vanishing gradient problem with a gated architecture: forget gate (what to discard), input gate (what to store), output gate (what to output), and a cell state that carries information across long sequences.

Related

• Vanilla RNN (simpler but limited)
• GRU (simplified LSTM variant)
• Transformers (largely replaced LSTMs for most tasks)

---

deep-learning rnn lstm