Research map / Label-efficient and robust learning / Recognition and long-tail
Analyzing and Improving Representations with the Soft Nearest Neighbor Loss
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 |
|---|---|---|---|
| Add a soft nearest neighbor loss bonus to cross-entropy training | CNN · MNIST test | accuracy (average): 98.82% -> 99.16% | Helped |
| Add a soft nearest neighbor loss bonus to cross-entropy training | CNN · Fashion-MNIST test | accuracy (average): 90.25% -> 91.06% | Helped |
| Add a soft nearest neighbor loss bonus to cross-entropy training | CNN · SVHN test | accuracy (average): 89.71% -> 89.90% | Helped |
| Add a soft nearest neighbor loss bonus to cross-entropy training | ResNet · CIFAR10 test | accuracy (average): 89.713% -> 89.900% | Helped |
| Maximize triplet loss instead of soft nearest neighbor loss as a regularizer (ablation) | MNIST classifier · MNIST adversarial examples | DkNN calibration | No clear effect |
| Replace the GAN discriminator with a pixel-space soft nearest neighbor objective (ablation) | GAN generator · MNIST generation | generation quality | Helped |
| Replace the GAN discriminator with a pixel-space soft nearest neighbor objective (ablation) | GAN generator · CIFAR10 generation | generation quality | Hurt |
| Add a soft nearest neighbor loss bonus to cross-entropy training | Classifier with DkNN · SVHN adversarial examples | correlation between DkNN credibility and accuracy | Helped |
| Add a soft nearest neighbor loss bonus to cross-entropy training | Classifier with DkNN · SVHN and CIFAR10 out-of-distribution inputs | out-of-distribution separation | Helped |
Closest papers by meaning
- The Effect of Manifold Entanglement and Intrinsic Dimensionality on Learning (AAAI 2022 · similarity 0.88)
- Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations (ICML 2019 · similarity 0.88)
- Disentangled Representation Learning with the Gromov-Monge Gap (ICLR 2025 · similarity 0.87)
- Reducing Class-wise Confusion for Incremental Learning with Disentangled Manifolds (CVPR 2025 · similarity 0.87)
- Disentanglement Learning via Topology (ICML 2024 · similarity 0.86)
- Stabilizing Sample Similarity in Representation via Mitigating Random Consistency (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.