AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

L2ight: Enabling On-Chip Learning for Optical Neural Networks via Efficient in-situ Subspace Optimization

Jiaqi Gu, Hanqing Zhu, Chenghao Feng, Zixuan Jiang, Ray Chen, David Z. Pan

NeurIPS 2021 · Architectures and efficiency › RNNs and normalization · approach: Scalable approximation and equivariant compute · 7 citations (OpenAlex)

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 changedWhereMeasuredEffect
Add feedback sampling (ablation)L2ight-SL · CIFAR-10, VGG-8normalized steps, total: 130.14 -> 73.89Improved
Add feedback sampling (ablation)L2ight-SL · CIFAR-10, ResNet-18accuracy (%): 92.37$_{\pm0.08}$ -> 91.35$_{\pm0.03}$Hurt
Add feedback sampling (ablation)L2ight-SL · CIFAR-10, ResNet-18normalized steps, total: 1,969.48 (1.00) -> 1,237.85 (1.59)Improved
Add column sampling after feedback sampling (ablation)L2ight-SL · CIFAR-10, VGG-8normalized PTC energy, total: 29.13 (1.17) -> 19.12 (1.78)Improved
Add column sampling after feedback sampling (ablation)L2ight-SL · CIFAR-10, VGG-8accuracy (%): 86.41$_{\pm0.25}$ -> 85.58$_{\pm0.01}$Hurt
Add data sampling after feedback and column sampling (ablation)L2ight-SL · CIFAR-10, VGG-8normalized steps, total: 73.07 (1.78) -> 36.54 (3.56)Improved
Add data sampling after feedback and column sampling (ablation)L2ight-SL · CIFAR-10, VGG-8accuracy (%): 85.58$_{\pm0.01}$ -> 84.45$_{\pm0.45}$Hurt
Use full calibration, mapping, and sparse subspace learning flow (ablation)L2ight-SL baseline · CIFAR-10, VGG-8accuracy (%): 86.66$_{\pm0.13}$ -> 90.20$_{\pm0.05}$Improved
Use full calibration, mapping, and sparse subspace learning flow (ablation)L2ight-SL baseline · CIFAR-10, VGG-8normalized PTC energy, total: 34.08 (1.00) -> 0.96 (35.64)Improved
Use spatial sampling (RAD) (ablation)L2ight-SL · CIFAR-10, VGG-8normalized steps, total: 130.14 (1.00) -> 130.14 (1.00)No clear effect
Use SWAT-U weight and spatial sampling (ablation)L2ight-SL · CIFAR-10, VGG-8accuracy (%): 86.66$_{\pm0.13}$ -> 73.91$_{\pm0.27}$Hurt
Use balanced top-K feedback sampling rather than uniform or top-K sampling (ablation)L2ight sparse subspace learning · feedback gradient approximationgradient angular similarityImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy