Research map / Label-efficient and robust learning / Self-supervised and contrastive
Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Loss
Jeff Z. HaoChen, Colin Wei, Adrien Gaidon, Tengyu Ma
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 |
|---|---|---|---|
| spectral contrastive loss instead of SimCLR/SimSiam | ResNet encoder · CIFAR-10/100, Tiny-ImageNet linear eval | top-1 accuracy: 90.60 -> 92.07 (CIFAR-10, 800 ep vs SimCLR) | Mixed |
| spectral contrastive loss vs SimSiam on ImageNet | ResNet 100-epoch · ImageNet linear eval | top-1 accuracy (%): 68.1 -> 66.97 | No clear effect |
| increase projection radius hyperparameter mu from 1 to 10 (ablation) | Spectral contrastive loss · Tiny-ImageNet 200 epochs | top-1 accuracy: 28.76 -> 41.30 | Improved |
| increase mu from 3 to 10 with longer training (ablation) | Spectral contrastive loss · CIFAR-100 800 epochs | top-1 accuracy: 66.18 -> 65.16 | Hurt |
| increase mu from 1 to 10 at short training (ablation) | Spectral contrastive loss · CIFAR-10 200 epochs | top-1 accuracy: 86.47 -> 88.66 | Improved |
Closest papers by meaning
- Zero-Mean Regularized Spectral Contrastive Learning: Implicitly Mitigating Wrong Connections in Positive-Pair Graphs (ICLR 2024 · similarity 0.91)
- Self-Supervised Contrastive Learning is Approximately Supervised Contrastive Learning (NeurIPS 2025 · similarity 0.89)
- An Augmentation-Aware Theory for Self-Supervised Contrastive Learning (ICML 2025 · similarity 0.88)
- Contrastive Learning is Spectral Clustering on Similarity Graph (ICLR 2024 · similarity 0.87)
- Perfect Alignment May be Poisonous to Graph Contrastive Learning (ICML 2024 · similarity 0.87)
- On the Similarities of Embeddings in Contrastive Learning (ICML 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.
