Research map / Label-efficient and robust learning / Recognition and long-tail
Contrastive Learning with Hard Negative Samples
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 hard negative sampling instead of uniform negatives | SimCLR · CIFAR100 | linear readout accuracy: +3% | Helped |
| Use hard negative sampling instead of uniform negatives | SimCLR · STL10 | linear readout accuracy: +7.3% | Helped |
| Use hard negative sampling instead of uniform negatives | SimCLR · tinyImageNet | Top-1 linear readout accuracy: +3.6% | Helped |
| Use hard negative sampling instead of uniform negatives | SimCLR · CIFAR10 | linear readout accuracy | Mixed |
| Use hard negative sampling instead of debiased sampling | SimCLR · CIFAR100 and STL10 | linear readout accuracy: +1.9% and +3.2% | Helped |
| Use hard negative sampling | InfoGraph · Graph classification (8 datasets) | accuracy: +3.2% on ENZYMES; +2.4% on REDDIT; +1-2% on DD and PTC; at least +0.5% on IMDB-B and MUTAG | Mixed |
| Use hard negative sampling with β=1 | Quick-thoughts · Sentence tasks (MR, CR, SUBJ, MPQA, TREC, MSRP) | accuracy/F1 | Mixed |
| Increase hardness from β=1 to β=2 (ablation) | Quick-thoughts hard sampling · CR | accuracy: 82.5 -> 83.6 | Helped |
| Increase hardness from β=1 to β=2 (ablation) | Quick-thoughts hard sampling · TREC | accuracy: 89.2 -> 88.7 | Hurt |
| Increase hardness from β=1 to β=2 (ablation) | Quick-thoughts hard sampling · Sentence tasks (MR, CR, SUBJ, MPQA, TREC, MSRP) | accuracy/F1: MR 77.1 -> 77.4; CR 82.5 -> 83.6; SUBJ 87.0 -> 86.8; MPQA 92.9 -> 93.4; TREC 89.2 -> 88.7; MSRP Acc 73.9 -> 73.5; MSRP F | Mixed |
| Use both hard sampling and debiasing rather than either alone (ablation) | SimCLR · STL10 | linear readout accuracy | Helped |
| Use hard sampling without debiasing instead of the standard baseline (ablation) | SimCLR · STL10 | linear readout accuracy | Helped |
Closest papers by meaning
- Difficulty-Based Sampling for Debiased Contrastive Representation Learning (CVPR 2023 · similarity 0.92)
- Difficult Examples Hurt Unsupervised Contrastive Learning: A Theoretical Perspective (ICLR 2026 · similarity 0.90)
- $\mathbb{X}$-Sample Contrastive Loss: Improving Contrastive Learning with Sample Similarity Graphs (ICLR 2025 · similarity 0.87)
- On the Similarities of Embeddings in Contrastive Learning (ICML 2025 · similarity 0.87)
- Self-Supervised Contrastive Learning is Approximately Supervised Contrastive Learning (NeurIPS 2025 · similarity 0.86)
- Optimal Sample Complexity of Contrastive Learning (ICLR 2024 · 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.