// ⬡ Deep Learning Foundations
Gated Recurrent Unit (GRU)
A leaner gated RNN than LSTM, with two gates for remembering and forgetting.
门控循环单元（GRU） 2014 Foundational core
IN PLAIN WORDS
GRU merges LSTM’s three gates into an update gate and a reset gate and drops the separate cell state. Fewer parameters, faster training, and on many sequence tasks it matches LSTM. It was introduced in the same paper as the RNN encoder–decoder.
Used in: Lightweight sequence modelling, time-series forecasting, early neural machine translation
01Original paper
Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
Kyunghyun Cho, Bart van Merrienboer, Caglar Gulcehre, et al. · 2014
arXiv ↗ ✓ Verified · arXiv API
02Where it sits
Learn the left column first; the right column builds on this one.
Learn first
Then you can learn
- Nothing in the atlas builds on it yet
Related:Sequence-to-Sequence LearningAttention (Bahdanau)
Left: prerequisites · right: builds on this · dashed: related. Full map →
03Learning resources
Official docs, explainers, courses and reference code. All links checked.
04Where the course teaches it
For AI / data
- Verification: verified, checked 2026-10-22.
- JSON: /api/learn/algorithms.json (id:
gru)