Data to 7 Oct 2026

Contrastive Learning with Hard Negative Samples

arXiv 2010.04592 ↗First posted 9 Oct 2020cs.LG stat.MLcited by 619

Not on the map. Why?

Cited by 29 in Multimodal LLMs / vision-language models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

How can you sample good negative examples for contrastive learning? We argue that, as with metric learning, contrastive learning of representations benefits from hard negative samples (i.e., points that are difficult to distinguish from an anchor point). The key challenge toward using hard negatives is that contrastive methods must remain unsupervised, making it infeasible to adopt existing negative sampling strategies that use true similarity information. In response, we develop a new family of unsupervised sampling methods for selecting hard negative samples where the user can control the hardness. A limiting case of this sampling results in a representation that tightly clusters each class, and pushes different classes as far apart as possible. The proposed method improves downstream performance across multiple modalities, requires only few additional lines of code to implement, and introduces no computational overhead.

Key facts. Cited by 619 papers in the map (data through 7 Oct 2026). First posted on arXiv 9 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 75; 2022: 112; 2023: 121; 2024: 128; 2025: 89; 2026: 94.

619 papers.

Sort:

By area:

2022When and why vision-language models behave like bags-of-words, and what to do about it?cited by 603
2024Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translationcited by 390
2022One Embedder, Any Task: Instruction-Finetuned Text Embeddingscited by 329
2021Graph Self-Supervised Learning: A Surveycited by 289
2023A Cookbook of Self-Supervised Learningcited by 248
2021Co2L: Contrastive Continual Learningcited by 243
2021Exploring Cross-Image Pixel Contrast for Semantic Segmentationcited by 242
2021Contrastive Learning Inverts the Data Generating Processcited by 220
2026Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Rankingcited by 211
2023OpenShape: Scaling Up 3D Shape Representation Towards Open-World Understandingcited by 183
2022SimGRACE: A Simple Framework for Graph Contrastive Learning without Data Augmentationcited by 171
2022Correct-N-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlationscited by 171

38 more

2024GME: Improving Universal Multimodal Retrieval by Multimodal LLMscited by 166
2021On Feature Decorrelation in Self-Supervised Learningcited by 160
2021Decoupled Contrastive Learningcited by 141
2021Self-supervised Learning on Graphs: Contrastive, Generative,or Predictivecited by 136
2022How to Exploit Hyperspherical Embeddings for Out-of-Distribution Detection?cited by 128
2021Neighborhood Contrastive Learning for Novel Class Discoverycited by 119
2021Can contrastive learning avoid shortcut solutions?cited by 114
2021An Empirical Study of Graph Contrastive Learningcited by 113
2021Dual Contrastive Learning for General Face Forgery Detectioncited by 100
2021Large-scale Unsupervised Semantic Segmentationcited by 98
2023Filtering, Distillation, and Hard Negatives for Vision-Language Pre-Trainingcited by 91
2023ARNIQA: Learning Distortion Manifold for Image Quality Assessmentcited by 87
2021Weakly Supervised Semantic Segmentation by Pixel-to-Prototype Contrastcited by 85
2024Perplexed by Perplexity: Perplexity-Based Data Pruning With Small Reference Modelscited by 83
2023Contrastive Preference Learning: Learning from Human Feedback without RLcited by 77
2023Verbs in Action: Improving verb understanding in video-language modelscited by 76
2025ReasonIR: Training Retrievers for Reasoning Taskscited by 76
2025Breaking the Modality Barrier: Universal Embedding Learning with Multimodal LLMscited by 73
2021CrossCLR: Cross-modal Contrastive Learning For Multi-modal Video Representationscited by 72
2021ProGCL: Rethinking Hard Negative Mining in Graph Contrastive Learningcited by 72
2022Plausible May Not Be Faithful: Probing Object Hallucination in Vision-Language Pre-trainingcited by 71
2022Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?cited by 67
2023Align and Attend: Multimodal Summarization with Dual Contrastive Lossescited by 65
2021Improving Text-to-Image Synthesis Using Contrastive Learningcited by 62
2022Contrastive Adapters for Foundation Model Group Robustnesscited by 62
2022Hard Sample Aware Network for Contrastive Deep Graph Clusteringcited by 61
2024EchoSight: Advancing Visual-Language Models with Wiki Knowledgecited by 61
2023Self-Supervised Multimodal Learning: A Surveycited by 58
2021Rebooting ACGAN: Auxiliary Classifier GANs with Stable Trainingcited by 57
2024TripletCLIP: Improving Compositional Reasoning of CLIP via Synthetic Vision-Language Negativescited by 57
2021Topology-Imbalance Learning for Semi-Supervised Node Classificationcited by 56
2025LLaVE: Large Language and Vision Embedding Models with Hardness-Weighted Contrastive Learningcited by 54
2021Self-Contrastive Learning with Hard Negative Sampling for Self-supervised Point Cloud Learningcited by 52
2022Dual Temperature Helps Contrastive Learning Without Many Negative Samples: Towards Understanding and Simplifying MoCocited by 52
2022Semi-Supervised and Unsupervised Deep Visual Learning: A Surveycited by 52
2023TimesURL: Self-supervised Contrastive Learning for Universal Time Series Representation Learningcited by 51
2023Magnushammer: A Transformer-Based Approach to Premise Selectioncited by 50
2021Contrastive Learning with Continuous Proxy Meta-Data for 3D MRI Classificationcited by 49

50 most cited of 619 All 619 (JSON)