Research map / Architectures and efficiency / RNNs and normalization
Learning to Remember Rare Events
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add a persistent nearest-neighbor memory module to a ConvNet | ConvNet · Omniglot, 5-way and 20-way one-shot and multi-shot tasks | accuracy | Helped |
| Add the memory module | Extended Neural GPU · Synthetic task | sequence accuracy: 12.2% -> 71.3% | Helped |
| Add the memory module | Sequence-to-Sequence with Attention · Synthetic task | sequence accuracy: 0.9% -> 35.2% | Helped |
| Add the memory module | GNMT · WMT14 En-De, Full Test | BLEU: 23.25 -> 23.29 | No clear effect |
| Update memory with even-numbered test lines before evaluating odd lines | GNMT with Memory Module · WMT14 En-De, Odd Test | BLEU: 23.16 -> 23.60 | Helped |
| Update memory with the whole test set before evaluation | GNMT with Memory Module · WMT14 En-De, Full Test | BLEU: 31.11 | Helped |
Closest papers by meaning
- ATLAS: Learning to Optimally Memorize the Context at Test Time (ICML 2026 · similarity 0.87)
- Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes (NeurIPS 2016 · similarity 0.87)
- Labeled Memory Networks for Online Model Adaptation (AAAI 2018 · similarity 0.87)
- Titans: Learning to Memorize at Test Time (NeurIPS 2025 · similarity 0.86)
- Do Your Best and Get Enough Rest for Continual Learning (CVPR 2025 · similarity 0.85)
- Towards flexible perception with visual memory (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.