diff --git a/backdoor/README.md b/backdoor/README.md
index 463622b..4638384 100644
--- a/backdoor/README.md
+++ b/backdoor/README.md
@@ -1,6 +1,6 @@
 # Backdoor in a MNIST CNN model
 
-A workflow to infect a PyTorch digit recognition CNN with a backdoor. Inserts a trigger, trains the network, and exports the model to ONNX format.
+A workflow to infect a [PyTorch digit recognition CNN](https://github.com/pytorch/examples/tree/master/mnist) with a [backdoor](https://arxiv.org/abs/1708.06733). Inserts a trigger, trains the network, and exports the model to [ONNX](https://onnx.ai/) format.
 
 Steps:
 1. MNIST dataset is downloaded from PyTorch repo
@@ -53,4 +53,4 @@ Which will launch a job to train the model on Peregrine using the GPU nodes.
 Inspired by:
 - [ShihaoZhaoZSH/BadNet](https://github.com/ShihaoZhaoZSH/BadNet)
 - [Kooscii/Badnets](https://github.com/Kooscii/BadNets)
-- [BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain (Gu et al, 2019)](https://arxiv.org/abs/1708.06733)
\ No newline at end of file
+- [BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain (Gu et al, 2019)](https://arxiv.org/abs/1708.06733)