Jailbreaking DeepSeek
Author(s): Shashwat (Shawn) Gupta Originally published on Towards AI. The AI community knows that DeepSeek censors some content. However, in this post, I bring out common approaches to jailbreak the model and get relevant information. The whole idea is to fool the …
Explainability for 3DResNet Classifier
Author(s): Shashwat (Shawn) Gupta Originally published on Towards AI. GradCAM is one of the simplest techniques to get explainability insights into model prediction. I was surprised to find that while there are many blogs on Medium about using GradCAM with ResNet, there …