Research map / Architectures and efficiency / RNNs and normalization
Dense Associative Memory for Pattern Recognition
Dmitry Krotov, John J. Hopfield
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Increase energy power from n=2 to n=3 | Dense associative memory · MNIST | test error: 1.6% | Improved |
| Increase energy power from n=2 to n=3 | Dense associative memory · MNIST | epochs to cross 2% error: 179-312 -> 158-262 | Improved |
| Use energy power n=2 or n=3 instead of n≥5 | Dense associative memory with N=100 and K=2000 · Random binary memory retrieval | perfect memory recovery: K=2000; K^max_no errors≈11 for n=2 and K^max_no errors≈360 for n=3 | Hurt |
| Use n=4 with K=2000 | Dense associative memory · Random binary memory retrieval, N=100 | overlap with closest memory: K^max_no errors≈7240 | Improved |
| Use higher rectified polynomial powers instead of ReLU (n=2) | Dense associative memory / equivalent one-hidden-layer network · MNIST | epochs to cross 2% error: 179-312 -> 158-262 | Improved |
| Use cross-entropy with softmax instead of the stated objective and output (ablation) | Dense associative memory classifier · MNIST | classification performance | Hurt |
| Increase the objective power m, emphasizing examples with largest target discrepancy (ablation) | Dense associative memory classifier · MNIST, at least for large n | generalization | Improved |
Closest papers by meaning
- Dense Associative Memory Through the Lens of Random Features (NeurIPS 2024 · similarity 0.88)
- Dense Associative Memory with Epanechnikov Energy (NeurIPS 2025 · similarity 0.87)
- Hyperparameter Transfer for Dense Associative Memories (NeurIPS 2026 · similarity 0.85)
- Semantically-correlated memories in a dense associative model (ICML 2024 · similarity 0.85)
- Learning Associative Memories with Gradient Descent (ICML 2024 · similarity 0.85)
- Factual recall in linear associative memories: sharp asymptotics and mechanistic insights (NeurIPS 2026 · 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.
