Dive into Deep Learning

By Aston Zhang, Zachary Chase Lipton, Mu Li, Alexander J. Smola. Journal of the American College of Radiology, 2020

682 citations recorded by the source corpus.

This open-source book represents our attempt to make deep learning approachable, teaching readers the concepts, the context, and the code. The entire book is drafted in Jupyter notebooks, seamlessly integrating exposition figures, math, and interactive examples with self-contained code. Our goal is to offer a resource that could (i) be freely available for everyone; (ii) offer sufficient technical depth to provide a starting point on the path to actually becoming an applied machine learning scientist; (iii) include runnable code, showing readers how to solve problems in practice; (iv) allow for rapid updates, both by us and also by the community at large; (v) be complemented by a forum for interactive discussion of technical details and to answer questions.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/bd17620c6cb5ca97ef773499223d1509d123745f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Understanding Continual Learning Settings with Data Distribution Drift Analysis (2021, background): “This type of drift is also known as “label shift” (Zhang et al., 2020).”
Explainable AI Methods - A Brief Overview (2020, background): “Machine learning (ML) and particularly the success of “deep learning” in the last decade made AI extremely popular again [15], [25], [90].”
A Survey of Deep Learning for Scientific Discovery (2020, methodology): “…of Toronto‘s Machine Learning Course or Stanford‘s CS229, detailed articles at https://towardsdatascience . com/ and even interactive textbooks such as https://d2l . ai/index . html (called Dive into Deep Learning [267]) and https://github . com/rasbt/python-machine-learning-book-2nd-edition .”
Beyond Fully-Connected Layers with Quaternions: Parameterization of Hypercomplex Multiplications with 1/n Parameters (2021, background): “The fully-connected layer is one of the most dominant components in existing deep learning literature (Goodfellow et al., 2016; Zhang et al., 2020).”
Is ChatGPT a General-Purpose Natural Language Processing Task Solver? (2023, background): “The key findings and insights are summarized as follows: • Although ChatGPT shows some capability as a generalist model that can perform multiple tasks (Zhang et al., 2021), it often performs worse than models that are fine-tuned on a given task (Section 4.3 and Figure 1).”
Transformers learn in-context by gradient descent (2022, methodology): “…choice of kernel function here and that this view still holds when m(xj) = 1 i.e. consider Transformers without MLPs or leverage the well-known view of softmax self-attention layer as a kernel function used to measure similarity between tokens (e.g. Choromanski et al., 2021; Zhang et al., 2021).”