Research map / Architectures and efficiency / RNNs and normalization
A New Initialization to Control Gradients in Sinusoidal Neural Networks
Andrea Combette, Nelly Pustelnik, Antoine Venaille
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 proposed SIREN init (c_w=sqrt3, c_b=0, sigma_a=0) instead of original Sitzmann init (ablation) | SIREN · image fitting and function approximation across depths | reconstruction noise / generalization | Improved |
| use sigma_a=0 instead of sigma_a=1 along sigma_g=1 curve (ablation) | SIREN proposed init · deep image fitting | output noise / spurious high frequencies | Improved |
| use proposed init instead of original SIREN init at large depth (ablation) | SIREN · NTK trace and input gradient vs depth | gradient norm / NTK eigenvalues: original grows exponentially (sigma_g about sqrt(1.2)); proposed sigma_g = 1 constant gradients | Improved |
| use PyTorch default init instead of proposed init (ablation) | SIREN · NTK trace and input gradient vs depth | gradient norm: sigma_g = sqrt(1/3) | Hurt |
| use proposed init (sigma_a=0) for Fourier spectrum control (ablation) | SIREN · 1D Fourier spectra, depths 4-32 | spectral energy beyond w0 | Improved |
Closest papers by meaning
- Sinusoidal Initialization, Time for a New Start (NeurIPS 2025 · similarity 0.91)
- Fast Training of Sinusoidal Neural Fields via Scaling Initialization (ICLR 2025 · similarity 0.89)
- VI^3NR: Variance Informed Initialization for Implicit Neural Representations (CVPR 2025 · similarity 0.88)
- From Activation to Initialization: Scaling Insights for Optimizing Neural Fields (CVPR 2024 · similarity 0.87)
- Tuning the Frequencies: Robust Training for Sinusoidal Neural Networks (CVPR 2025 · similarity 0.87)
- Robust Weight Initialization for Tanh Neural Networks with Fixed Point Analysis (ICLR 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.
