# karpathy/micrograd architecture

> Architecture diagram of the GitHub repository karpathy/micrograd, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-19). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/karpathy/micrograd
- Repository: https://github.com/karpathy/micrograd

## Overview

micrograd is a small scalar reverse-mode autodiff library with a PyTorch-like neural-network layer. A user builds a computation from scalar Values, calls backward to obtain gradients, and can use the neural-network API in a documented classifier-training demo. A separate notebook demonstrates graph visualization. The core engine and neural-network code were inspected; demo and visualization behavior is represented only to the extent described in the README.

Main entry points: [`demo.ipynb`](https://github.com/karpathy/micrograd/blob/master/demo.ipynb) (Classifier Demo), [`micrograd/nn.py`](https://github.com/karpathy/micrograd/blob/master/micrograd/nn.py) (Network Modules), [`micrograd/engine.py`](https://github.com/karpathy/micrograd/blob/master/micrograd/engine.py) (Value Engine), [`trace_graph.ipynb`](https://github.com/karpathy/micrograd/blob/master/trace_graph.ipynb) (Graph Visualization)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_autodiff["Scalar Autodiff"]
  node_engine["Value Engine<br/>[engine.py]"]
  node_graph_demo["Graph Visualization<br/>[trace_graph.ipynb]"]
end

subgraph group_neural["Neural Networks"]
  node_nn["Network Modules<br/>[nn.py]"]
  node_classifier_demo["Classifier Demo<br/>[demo.ipynb]"]
  node_classifier_result["Decision Boundary<br/>[demo.ipynb]"]
end

node_user(("Library User"))

node_user -->|"builds values"| node_engine
node_user -->|"runs backward"| node_engine
node_nn -->|"uses values"| node_engine
node_user -->|"trains with"| node_classifier_demo
node_classifier_demo -->|"initializes network"| node_nn
node_classifier_demo -->|"produces"| node_classifier_result
node_user -->|"visualizes graph"| node_graph_demo
node_graph_demo -->|"demonstrates neuron"| node_nn

click node_engine "https://github.com/karpathy/micrograd/blob/master/micrograd/engine.py"
click node_nn "https://github.com/karpathy/micrograd/blob/master/micrograd/nn.py"
click node_classifier_demo "https://github.com/karpathy/micrograd/blob/master/demo.ipynb"
click node_graph_demo "https://github.com/karpathy/micrograd/blob/master/trace_graph.ipynb"
click node_classifier_result "https://github.com/karpathy/micrograd/blob/master/demo.ipynb"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_engine,node_graph_demo,node_user toneBlue
class node_nn,node_classifier_demo,node_classifier_result toneAmber
```

## Components

### Scalar Autodiff

- [Value Engine](https://github.com/karpathy/micrograd/blob/master/micrograd/engine.py): `micrograd/engine.py`
- [Graph Visualization](https://github.com/karpathy/micrograd/blob/master/trace_graph.ipynb): `trace_graph.ipynb`

### Neural Networks

- [Network Modules](https://github.com/karpathy/micrograd/blob/master/micrograd/nn.py): `micrograd/nn.py`
- [Classifier Demo](https://github.com/karpathy/micrograd/blob/master/demo.ipynb): `demo.ipynb`
- [Decision Boundary](https://github.com/karpathy/micrograd/blob/master/demo.ipynb): `demo.ipynb`

### Other components

- **Library User**

## Connections

- Library User → Value Engine: builds values (evidence: `README.md`)
- Library User → Value Engine: runs backward (evidence: `README.md`)
- Network Modules → Value Engine: uses values (evidence: `micrograd/nn.py`)
- Library User → Classifier Demo: trains with (evidence: `README.md`)
- Classifier Demo → Network Modules: initializes network (evidence: `README.md`)
- Classifier Demo → Decision Boundary: produces (evidence: `README.md`)
- Library User → Graph Visualization: visualizes graph (evidence: `README.md`)
- Graph Visualization → Network Modules: demonstrates neuron (evidence: `README.md`)
