Unsupervised High-level Feature Learner

Closed weights Google 1B parameters July 2012

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Google
Organisation type
Industry
Country
United States of America
Published
12 July 2012
Authors
Quoc V. Le, Marc'Aurelio Ranzato, Rajat Monga, Matthieu Devin, Kai Chen, Greg S. Corrado, Jeff Dean, Andrew Y. Ng

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Vision
Task
Image classification
Approach
Unsupervised

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
1B

"To answer this, we train a 9-layered locally connected sparse autoencoder with pooling and local contrast normalization on a large dataset of images (the model has 1 billion connections, the dataset has 10 million 200x200 pixel images downloaded from the Internet)"

Training data
1,200,000,000,000 tokens

10 million 200x200 images extracted from Youtube videos

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute
6 × 10¹⁷ FLOP

Assuming 1 epoch, 10 million images and 1 billion parameters, 6*N*D = 6*10^17 FLOP

How it was established
Operation counting

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Wall-clock time
72 hours

"We train this network using model parallelism and asynchronous SGD on a cluster with 1,000 machines (16,000 cores) for three days. "

Compute cost
$16

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Frontier model
Yes
Why it is tracked
Highly cited,SOTA improvement

"we trained our network to obtain 15.8% accuracy in recognizing 20,000 object categories from ImageNet, a leap of 70% relative improvement over the previous state-of-the-art"

Record confidence
Likely
Citations
2,909

Sources

Where this record came from and when it was last checked.

Reference
Building High-level Features Using Large Scale Unsupervised Learning
Last updated
28 November 2025

What the numbers mean

About this model

Unsupervised High-level Feature Learner was published by Google, in United States of America, in July 2012. industry is the category the publisher falls under.

It works in Vision, and is recorded as doing image classification.

Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.

Training and provenance

Training it took roughly 6 × 10¹⁷ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.

The training set ran to roughly 1,200,000,000,000 tokens.

Its inclusion criterion is highly cited,SOTA improvement.

Answers

Unsupervised High-level Feature Learner — common questions

01

What is Unsupervised High-level Feature Learner used for?

Unsupervised High-level Feature Learner works in Vision, and is recorded as handling image classification. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

02

How much compute was used to train Unsupervised High-level Feature Learner?

Around 6 × 10¹⁷ FLOP. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

03

What GPU do I need to run Unsupervised High-level Feature Learner?

None. Unsupervised High-level Feature Learner is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

04

Is Unsupervised High-level Feature Learner open source?

The licensing for Unsupervised High-level Feature Learner was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.

05

How many parameters does Unsupervised High-level Feature Learner have?

Unsupervised High-level Feature Learner has 1B parameters. "To answer this, we train a 9-layered locally connected sparse autoencoder with pooling and local contrast normalization on a large dataset of images (the model has 1 billion connections, the dataset has 10 million 200x200 pixel images downloaded from the Internet)". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

06

Who created Unsupervised High-level Feature Learner?

Unsupervised High-level Feature Learner was published by Google, based in United States of America, categorised as industry.

07

When was Unsupervised High-level Feature Learner released?

Unsupervised High-level Feature Learner was published in July 2012. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

Source

Original publication

Record last updated 28 November 2025

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.