Mathematical Expression

Batching speeds up training and inference, but for LLMs we can't just use matrices for it -- we need higher-order tensors.

In the second part of this two-part series, we'll look at more advanced Dialyzer use cases.

Why a mathematician thought this infinite series explained how God created the universe

A team of scientists has developed a novel hybrid scheme for both steady and unsteady single-phase compressible flow simulations. Their scheme has potential applications in real-world scenarios, offering a promising avenue for future research.

The busy beaver function is unpredictable. But now, after more than 40 years, the fifth value of the function has been revealed

Bézier curves are widely used for defining vector graphics. They are basically polynomial parametric curves, but given in the Bernstein basis, which enables us to define the curve using control points.

A blog about math by Nic Ford

SecureLoop is an MIT-developed search engine that can identify an optimal design for a deep neural network accelerator that preserves data security while improving energy efficiency and boosting performance. This could enable device manufacturers to increase the speed of demanding AI applications, while ensuring sensitive data remain safe from attackers.

How to use matrix operations from linear algebra in JavaScript. What are inverse, transpose and identity matrices and how can they help in machine learning ...

Article URL: https://realtimecollisiondetection.net/blog/?p=69 Comments URL: https://news.ycombinator.com/item?id=37258753 Points: 1 # Comments: 0

Get up and running with large language models, locally.

Math can explain why your laces spontaneously come untied — and how to stop it.

In Numpy, the `r_` and `c_` class objects are utilities that allow users to concatenate objects using slice notation. In this post, I explain how to make the most out of them by modifying its string notation and I measure their performance for concatenating and for reshaping.

TL;DR

This paper shows how to build a sparse tensor algebra compiler that is agnostic to tensor formats (data layouts). We develop an interface that describes formats in terms of their capabilities and properties, and show how to build a modular code generator where new formats can be added as plugins. We then describe six implementations of the interface that compose to form the dense, CSR/CSF, COO, DIA, ELL, and HASH…

This is the second article in that series, and is associated with our Intro to Deep Learning Github repository where you can find practical examples of many deep learning applications and tactics…

Article URL: https://plato.stanford.edu/entries/logic-combinatory/ Comments URL: https://news.ycombinator.com/item?id=35941979 Points: 1 # Comments: 0

Article URL: https://web.archive.org/web/20150222232543/http://www.the-idea-shop.com/article/216/deriving-the-poisson-distribution-from-the-binomial Comments URL: https://news.ycombinator.com/item?id=35769181 Points: 1 # Comments: 0

Let's look at how we can solve the quadratic equation without using the memorized formula, and then derive the quadratic formula itself from scratch using our method of …

In this article we will work with LPeg to parse basic arithmetic expressions. This is the first part of a series of articles for creating your own programming language.