Research map / Optimization / Implicit bias and generalization
On Single-Index Models beyond Gaussian Data
Aaron Zweig, Loucas Pillaud-Vivien, Joan Bruna
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 the non-monotonic link with information exponent s = 2 instead of the degree-three Hermite link with s = 3 | SGD on Gaussian and hypercube inputs · Gaussian and hypercube inputs | strong recovery | Mixed |
| Initialize above the largest-zero threshold rather than uniformly at random | SGD with the 4th-degree Gegenbauer link on spherical inputs · Uniform sphere, 4th-degree Gegenbauer link | strong recovery | Improved |
| Use random initialization instead of initializing above the largest-zero threshold | SGD with the 4th-degree Gegenbauer link on spherical inputs · Uniform sphere, 4th-degree Gegenbauer link | strong recovery | Hurt |
Closest papers by meaning
- Learning Orthogonal Multi-Index Models: A Fine-Grained Information Exponent Analysis (NeurIPS 2025 · similarity 0.91)
- Neural network learns low-dimensional polynomials with SGD near the information-theoretic limit (NeurIPS 2024 · similarity 0.89)
- Symmetric Single Index Learning (ICLR 2024 · similarity 0.89)
- Learning single index models via harmonic decomposition (NeurIPS 2025 · similarity 0.88)
- Sample and Computationally Efficient Robust Learning of Gaussian Single-Index Models (NeurIPS 2024 · similarity 0.88)
- Can Neural Networks Achieve Optimal Computational-statistical Tradeoff? An Analysis on Single-Index Model (ICLR 2025 · similarity 0.87)
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.
