Skip to main content

> ML_ALGORITHM // BOOTSTRAP-YOUR-OWN-LATENT-BYOL_v1.0

Bootstrap Your Own Latent (BYOL)

Self-supervised visual representation algorithm that achieves state-of-the-art representations without negative pairs by iteratively bootstrapping target network predictions.

Non-contrastive Self-Supervised Learningself-supervisedblack-boxlarge (>100k)
Back to All Algorithms
Computational Complexity
Training Complexity:O(epochs * batch_size * forward_pass)
Inference Complexity:O(forward_pass)
Hardware Profile
CPU Friendly:No
Requires GPU:Yes
Memory Footprint:moderate
Interpretability & Data
Interpretability Tier:black-box
Training Data Needs:large (>100k)

Interpretability Assessment

Generates robust visual representations without relying on negative samples or large contrastive batches.

Suitable Tasks & Supported Modalities

Suitable Tasks:
feature extractionimage classification
Supported Modalities:
image

Implementing Libraries

PyTorchLinux Foundation / PyTorch Foundation · v2.4.1
View Spec
torchvision

Foundational Literature

Common Pitfalls & Warnings
  • Removing the additional online predictor layer collapses all outputs into a trivial constant vector