PyTorch implementation for " Conditional Negative Sampling for Contrastive Learning of Visual Representations" (https://arxiv.org/abs/2010.02037).
-
Updated
Sep 15, 2021 - Python
PyTorch implementation for " Conditional Negative Sampling for Contrastive Learning of Visual Representations" (https://arxiv.org/abs/2010.02037).
Experimental code for our paper on informative and diverse sampling of negative examples for dense retrieval
Code for "Bypassing Skip-Gram Negative Sampling: Dimension Regularization as a More Efficient Alternative for Graph Embeddings" published in KDD'25
To associate your repository with the negative-sampling topic, visit your repo's landing page and select "manage topics."