{"url":"/method/byol","slug":"byol","name":"BYOL","full_name":"Bootstrap Your Own Latent","full_name_withheld":false,"description_markdown":"BYOL (Bootstrap Your Own Latent) is a new approach to self-supervised learning. BYOL’s goal is to learn a representation $y_θ$ which can then be used for downstream tasks. BYOL uses two neural networks to learn: the online and target networks. The online network is defined by a set of weights $θ$ and is comprised of three stages: an encoder $f_θ$, a projector $g_θ$ and a predictor $q_θ$. The target network has the same architecture\r\nas the online network, but uses a different set of weights $ξ$. The target network provides the regression\r\ntargets to train the online network, and its parameters $ξ$ are an exponential moving average of the\r\nonline parameters $θ$.\r\n\r\nGiven the architecture diagram on the right, BYOL minimizes a similarity loss between $q_θ(z_θ)$ and $sg(z'{_ξ})$, where $θ$ are the trained weights, $ξ$ are an exponential moving average of $θ$ and $sg$ means stop-gradient. At the end of training, everything but $f_θ$ is discarded, and $y_θ$ is used as the image representation.\r\n\r\nSource: [Bootstrap Your Own Latent - A New Approach to Self-Supervised Learning](https://paperswithcode.com/paper/bootstrap-your-own-latent-a-new-approach-to-1)\r\n\r\nImage credit: [Bootstrap Your Own Latent - A New Approach to Self-Supervised Learning](https://paperswithcode.com/paper/bootstrap-your-own-latent-a-new-approach-to-1)","description_state":"present","introduced_year":null,"introduced_by":{"title":null,"paper":null,"first_author":null,"n_authors":0,"url_abs":null,"archive_paper_url":null},"source":{"url":"http://proceedings.neurips.cc/paper/2020/hash/f3ada80d5c4ee70142b17b8192b2958e-Abstract.html","title":"Bootstrap Your Own Latent - A New Approach to Self-Supervised Learning","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"General","area_id":"general","collection":"Self-Supervised Learning","url":"/methods/category/self-supervised-learning","pwc_aliases":[]}],"n_papers_tagged":123,"archive_num_papers":null,"papers_newest_first":[{"paper":null,"title":"Representation Learning via Non-Contrastive Mutual Information","date":"2025-04-23","arxiv_id":"2504.16667","n_code_links":0,"syntology":null},{"paper":null,"title":"Variational Self-Supervised Learning","date":"2025-04-06","arxiv_id":"2504.04318","n_code_links":0,"syntology":null},{"paper":"/paper/implicit-contrastive-representation-learning-1","title":"Implicit Contrastive Representation Learning with Guided Stop-gradient","date":"2025-03-12","arxiv_id":"2503.09058","n_code_links":1,"syntology":{"ran":1,"of":1,"unverified":0,"pointer_only":1}},{"paper":"/paper/adversarial-network-embedding-with","title":"Adversarial network embedding with bootstrapped representations for sparse networks","date":"2025-03-03","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/dataset-ownership-verification-in-contrastive","title":"Dataset Ownership Verification in Contrastive Pre-trained Models","date":"2025-02-11","arxiv_id":"2502.07276","n_code_links":1,"syntology":null},{"paper":"/paper/enhancing-contrastive-learning-inspired-by","title":"Enhancing Contrastive Learning Inspired by the Philosophy of \"The Blind Men and the Elephant\"","date":"2024-12-21","arxiv_id":"2412.16522","n_code_links":1,"syntology":null},{"paper":null,"title":"Self-Supervised Radiograph Anatomical Region Classification -- How Clean Is Your Real-World Data?","date":"2024-12-20","arxiv_id":"2412.15967","n_code_links":0,"syntology":null},{"paper":null,"title":"Maximising Histopathology Segmentation using Minimal Labels via Self-Supervision","date":"2024-12-19","arxiv_id":"2412.15389","n_code_links":0,"syntology":null},{"paper":null,"title":"Bridge: A Unified Framework to Knowledge Graph Completion via Language Models and Knowledge Representation","date":"2024-11-11","arxiv_id":"2411.06660","n_code_links":0,"syntology":null},{"paper":null,"title":"Self-Supervised Anomaly Detection in the Wild: Favor Joint Embeddings Methods","date":"2024-10-05","arxiv_id":"2410.04289","n_code_links":0,"syntology":null},{"paper":"/paper/cross-model-cross-stream-learning-for-self","title":"Cross-Model Cross-Stream Learning for Self-Supervised Human Action Recognition","date":"2024-09-23","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/self-supervised-syllable-discovery-based-on","title":"Self-Supervised Syllable Discovery Based on Speaker-Disentangled HuBERT","date":"2024-09-16","arxiv_id":"2409.10103","n_code_links":1,"syntology":null},{"paper":null,"title":"Unfolding Videos Dynamics via Taylor Expansion","date":"2024-09-04","arxiv_id":"2409.02371","n_code_links":0,"syntology":null},{"paper":"/paper/cnn-jepa-self-supervised-pretraining","title":"CNN-JEPA: Self-Supervised Pretraining Convolutional Neural Networks Using Joint Embedding Predictive Architecture","date":"2024-08-14","arxiv_id":"2408.07514","n_code_links":1,"syntology":null},{"paper":"/paper/efficiency-for-free-ideal-data-are","title":"Efficiency for Free: Ideal Data Are Transportable Representations","date":"2024-05-23","arxiv_id":"2405.14669","n_code_links":1,"syntology":null},{"paper":"/paper/self-supervised-learning-for-classifying","title":"Self-supervised learning for classifying paranasal anomalies in the maxillary sinus","date":"2024-04-29","arxiv_id":"2404.18599","n_code_links":1,"syntology":null},{"paper":null,"title":"Integration of Self-Supervised BYOL in Semi-Supervised Medical Image Recognition","date":"2024-04-16","arxiv_id":"2404.10405","n_code_links":0,"syntology":null},{"paper":"/paper/masked-modeling-duo-towards-a-universal-audio","title":"Masked Modeling Duo: Towards a Universal Audio Pre-training Framework","date":"2024-04-09","arxiv_id":"2404.06095","n_code_links":2,"syntology":{"ran":7,"of":7,"unverified":0,"pointer_only":7}},{"paper":"/paper/self-supervised-visualisation-of-medical","title":"Self-supervised Visualisation of Medical Image Datasets","date":"2024-02-22","arxiv_id":"2402.14566","n_code_links":1,"syntology":null},{"paper":"/paper/the-common-stability-mechanism-behind-most","title":"The Common Stability Mechanism behind most Self-Supervised Learning Approaches","date":"2024-02-22","arxiv_id":"2402.14957","n_code_links":1,"syntology":null},{"paper":null,"title":"WERank: Towards Rank Degradation Prevention for Self-Supervised Learning Using Weight Regularization","date":"2024-02-14","arxiv_id":"2402.09586","n_code_links":0,"syntology":null},{"paper":"/paper/gps-ssl-guided-positive-sampling-to-inject","title":"GPS-SSL: Guided Positive Sampling to Inject Prior Into Self-Supervised Learning","date":"2024-01-03","arxiv_id":"2401.01990","n_code_links":1,"syntology":null},{"paper":null,"title":"Bootstrap Your Own Variance","date":"2023-12-06","arxiv_id":"2312.03213","n_code_links":0,"syntology":null},{"paper":null,"title":"SASSL: Enhancing Self-Supervised Learning via Neural Style Transfer","date":"2023-12-02","arxiv_id":"2312.01187","n_code_links":0,"syntology":null},{"paper":"/paper/cucl-codebook-for-unsupervised-continual","title":"CUCL: Codebook for Unsupervised Continual Learning","date":"2023-11-25","arxiv_id":"2311.14911","n_code_links":1,"syntology":null},{"paper":null,"title":"Feature Normalization Prevents Collapse of Non-contrastive Learning Dynamics","date":"2023-09-28","arxiv_id":"2309.16109","n_code_links":0,"syntology":null},{"paper":null,"title":"Adapting Self-Supervised Representations to Multi-Domain Setups","date":"2023-09-07","arxiv_id":"2309.03999","n_code_links":0,"syntology":null},{"paper":"/paper/feature-suppressed-contrast-for-self","title":"Feature-Suppressed Contrast for Self-Supervised Food Pre-training","date":"2023-08-07","arxiv_id":"2308.03272","n_code_links":1,"syntology":null},{"paper":"/paper/visual-geo-localization-with-self-supervised","title":"VG-SSL: Benchmarking Self-supervised Representation Learning Approaches for Visual Geo-localization","date":"2023-07-31","arxiv_id":"2308.00090","n_code_links":1,"syntology":null},{"paper":null,"title":"Self-Supervised Learning for Improved Synthetic Aperture Sonar Target Recognition","date":"2023-07-27","arxiv_id":"2307.15098","n_code_links":0,"syntology":null}],"papers_shown":30,"tasks":[{"task":"/task/self-supervised-learning","name":"Self-Supervised Learning","papers":79},{"task":"/task/representation-learning","name":"Representation Learning","papers":53},{"task":"/task/contrastive-learning","name":"Contrastive Learning","papers":36},{"task":"/task/linear-evaluation","name":"Linear evaluation","papers":15},{"task":"/task/data-augmentation","name":"Data Augmentation","papers":14},{"task":"/task/image-classification","name":"Image Classification","papers":11},{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":7},{"task":"/task/object-detection","name":"Object Detection","papers":6},{"task":"/task/transfer-learning","name":"Transfer Learning","papers":6},{"task":"/task/image-classification","name":"image-classification","papers":6},{"task":"/task/object-detection-1","name":"object-detection","papers":6},{"task":"/task/diversity","name":"Diversity","papers":5},{"task":"/task/action-recognition-in-videos","name":"Action Recognition","papers":4},{"task":"/task/clustering","name":"Clustering","papers":4},{"task":"/task/pseudo-label","name":"Pseudo Label","papers":4},{"task":"/task/time-series","name":"Time Series Analysis","papers":4},{"task":"/task/depth-estimation","name":"Depth Estimation","papers":3},{"task":"/task/disentanglement","name":"Disentanglement","papers":3},{"task":"/task/few-shot-learning","name":"Few-Shot Learning","papers":3},{"task":"/task/segmentation","name":"Segmentation","papers":3}],"tasks_shown":20,"n_tasks":124,"usage_by_year":[{"year":"2020","papers":3},{"year":"2021","papers":39},{"year":"2022","papers":36},{"year":"2023","papers":23},{"year":"2024","papers":17},{"year":"2025","papers":5}],"row_source":"embedded","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/byol"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}