Research map / Architectures and efficiency / Attention and transformers
What can a Single Attention Layer Learn? A Study Through the Random Features Lens
Hengyu Fu, Tianyu Guo, Yu Bai, Song Mei
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 |
|---|---|---|---|
| Use random-feature attention (RFA) instead of a random-feature two-layer MLP (RFMLP) with equal parameter count | RFMLP · Simulated f_{1,2}, f_{1,4}, f_{2,3} targets, d=16, N=16 | test risk | Improved |
| Use unbiased RFA instead of biased RFA (identity-mean query-key matrix) on f_{1,2} and f_{1,4} | BRFA · Simulated f_{1,2}, f_{1,4} | test risk | Improved |
| Use biased RFA (non-zero-mean query-key weights) instead of zero-mean RFA on correlation-weighted target f_{2,3} | RFA · Simulated f_{2,3} | test risk | Improved |
Closest papers by meaning
- Single-Head Attention in High Dimensions: A Theory of Generalization, Weights Spectra, and Scaling Laws (ICML 2026 · similarity 0.89)
- Fundamental limits of learning in sequence multi-index models and deep attention networks: high-dimensional asymptotics and sharp thresholds (ICML 2025 · similarity 0.89)
- Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency (NeurIPS 2025 · similarity 0.88)
- Bayes optimal learning of attention-indexed models (NeurIPS 2025 · similarity 0.88)
- Attention layers provably solve single-location regression (ICLR 2025 · similarity 0.87)
- Mind the Gap: a Spectral Analysis of Rank Collapse and Signal Propagation in Attention Layers (ICML 2025 · 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.
