---
title: "Tensor Broadcasting"
description: "Add a per-feature offset to a (2, 3) array without copying it: match shapes from the right, then catch a broadcast that runs but computes the wrong thing."
canonical_url: "https://fanout.sh/ai/lessons/core-ai-intuitions-tensor-broadcasting"
md_url: "https://fanout.sh/ai/lessons/core-ai-intuitions-tensor-broadcasting.md"
access: "pro-preview"
---

# Tensor Broadcasting

Add a per-feature offset to a (2, 3) array without copying it: match shapes from the right, then catch a broadcast that runs but computes the wrong thing.

- Access: Fanout Pro (public preview)
- Module: Core AI Intuitions

## A glimpse of Tensor Broadcasting

You have two rows of measurements, with three features in each row. You want to add 10 to the first feature, 20 to the second and 30 to the third, for both rows.

Write the measurements as an array with shape (2, 3). Shape lists the length of each axis: two rows, three columns. In machine-learning code, arrays like these are called tensors.

## Continue learning

Pro unlocks guided learning, complete notes, practical walkthroughs, exercises, and every AI Research lesson.

---
This representation contains public Fanout content only. Protected Pro lessons, account data, billing, checkout, and pricing are not included.

Browse the public content map: https://fanout.sh/sitemap.md