Research map / Architectures and efficiency / RNNs and normalization
Large Associative Memory Problem in Neurobiology and Machine Learning
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 |
|---|---|---|---|
| Unfold the feature-only associative-memory model by adding hidden memory neurons with pairwise synapses | Dense Associative Memory · Associative memory | storage capacity | Helped |
| Use power-function hidden-neuron activation F(x)=x^n | Model A · Associative memory | storage capacity: N_{\text{mem}}\sim \min(N_f^{n-1}, N_h) | Helped |
| Use exponential hidden-neuron activation F(x)=\exp(x) | Model A · Associative memory | storage capacity: exponential storage capacity | Helped |
| Use quadratic hidden-neuron activation, corresponding to power n=2 | Model A · Associative memory | model equivalence | No clear effect |
| Use softmax contrastive normalization in the hidden layer and quadratic feature Lagrangian | General two-layer network · Associative memory | model equivalence | No clear effect |
| Use spherical normalization in the feature layer | General two-layer network · Spherical Memory | model formulation | No clear effect |
Closest papers by meaning
- Dense Associative Memory Through the Lens of Random Features (NeurIPS 2024 · similarity 0.91)
- Associative memory and dead neurons (ICLR 2025 · similarity 0.88)
- Exponential Dynamic Energy Network for High Capacity Sequence Memory (NeurIPS 2025 · similarity 0.87)
- Factual recall in linear associative memories: sharp asymptotics and mechanistic insights (NeurIPS 2026 · similarity 0.86)
- Relating Hopfield Networks to Episodic Control (NeurIPS 2024 · similarity 0.86)
- Semantically-correlated memories in a dense associative model (ICML 2024 · similarity 0.86)
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.