Research map / Learning theory and algorithms / Prediction and decision losses
Robust Implicit Networks via Non-Euclidean Contractions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 NEMON instead of implicit deep learning | MNIST image classification · MNIST | best accuracy: 0.9721 -> 0.9772 | Helped |
| Use NEMON instead of MON | MNIST image classification · MNIST | best accuracy: 0.9762 -> 0.9772 | Helped |
| Add input-output Lipschitz regularization (ablation) | NEMON · MNIST | input-output Lipschitz constant: two order of magnitude decrease | Helped |
| Increase the input-output Lipschitz regularization parameter (ablation) | NEMON · MNIST continuous image inversion attack | robustness | Helped |
| Use NEMON instead of implicit deep learning or MON | Image classification under continuous image inversion attack · MNIST | robustness | Helped |
| Use NEMON instead of MON | Image classification under ℓ∞-bounded perturbations · CIFAR-10 | robustness | Helped |
Closest papers by meaning
- Subhomogeneous Deep Equilibrium Models (ICML 2024 · similarity 0.90)
- Positive Concave Deep Equilibrium Models (ICML 2024 · similarity 0.87)
- DeltaDEQ: Exploiting Heterogeneous Convergence for Accelerating Deep Equilibrium Iterations (NeurIPS 2024 · similarity 0.86)
- Implicit regularization of deep residual networks towards neural ODEs (ICLR 2024 · similarity 0.86)
- Separation and Bias of Deep Equilibrium Models on Expressivity and Learning Dynamics (NeurIPS 2024 · similarity 0.86)
- Understanding Representation of Deep Equilibrium Models from Neural Collapse Perspective (NeurIPS 2024 · 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.
