Crossword

3. (Across) The phenomenon where a generative model confidently produces output that is factually incorrect or nonsensical.
4. (Across) A technique that converts high-dimensional categorical data (like words) into low-dimensional, continuous vectors.
7. (Down) The end-to-end workflow that automates the flow of data from raw input to a trained model.

1. (Down) A vector representing the direction and magnitude of the steepest increase of the loss function; crucial for the backpropagation step.
2. (Down) The formal structure or "blueprint" that defines how data is organized in a database.
5. (Down) A common regularization technique where randomly selected neurons are ignored during training to prevent overfitting.
6. (Across) The fundamental data structure in Deep Learning; essentially a multi-dimensional array or matrix.