Two Stage Feature Extraction (MNIST)

Closed weights New York University (NYU) 258.8K parameters September 2009

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
New York University (NYU)
Organisation type
Academia
Country
United States of America
Published
1 September 2009
Authors
Kevin Jarrett, K. Kavukcuoglu, Marc'Aurelio Ranzato, Yann LeCun

What it does

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

Domain
Vision
Task
Image classification

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
258.8K

CNN: 32*5*5+32*64*5*5=52000 FC: 1024*200+200*10=206800 Total: 52000+206800=258800

Training data
50,000 tokens

"experiments were run on the MNIST dataset, which contains 60,000 gray-scale 28x28 pixel digit images for training and 10,000 images for testing"

Epochs
33

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
2.1 × 10¹³ FLOP

Assuming no padding and stride 1 “on the 50,000 non-validation training samples until the best error rate on the validation set was reached (this took 30 epochs)“ First CNN layer: 2*32*24*24*5*5=921600 Second CNN layer: 2*64*16*5*5*8*8=3276800 First FC layer: 2*4*4*64*200=409600 Second FC layer: 2*200*10=4000 Total forward FLOP: 921600+3276800+409600+4000=4612000 Totral training compute: 4612000*3*30*50000=20754000000000 They train another model in this paper that I think is at least an order o…

How it was established
Operation counting

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased

How it is classified

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

Why it is tracked
Highly cited,SOTA improvement

They claim SOTA performance on the MNIST and NORB datasets. Is also highly cited

Record confidence
Confident

Sources

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

Reference
What is the best multi-stage architecture for object recognition?
Last updated
11 February 2026

What the numbers mean

About this model

Two Stage Feature Extraction (MNIST) was published by New York University (NYU), in United States of America, in September 2009. academia 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.

How it was trained

The training run consumed about 2.1 × 10¹³ FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

Around 50,000 tokens went into training it.

The reason it appears in this catalogue at all is highly cited,SOTA improvement.

Answers

Two Stage Feature Extraction (MNIST) — common questions

01

Is Two Stage Feature Extraction (MNIST) open source?

No. Two Stage Feature Extraction (MNIST) has not had its weights published, so it exists only as a service controlled by its owner.

02

How many parameters does Two Stage Feature Extraction (MNIST) have?

Two Stage Feature Extraction (MNIST) has 258.8K parameters. CNN: 32*5*5+32*64*5*5=52000 FC: 1024*200+200*10=206800 Total: 52000+206800=258800. 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.

03

Who created Two Stage Feature Extraction (MNIST)?

Two Stage Feature Extraction (MNIST) was published by New York University (NYU), based in United States of America, categorised as academia.

04

When was Two Stage Feature Extraction (MNIST) released?

Two Stage Feature Extraction (MNIST) was published in September 2009. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

05

What is Two Stage Feature Extraction (MNIST) used for?

Two Stage Feature Extraction (MNIST) works in Vision, and is recorded as handling image classification. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

06

How much compute was used to train Two Stage Feature Extraction (MNIST)?

Around 2.1 × 10¹³ 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.

07

What GPU do I need to run Two Stage Feature Extraction (MNIST)?

None. Two Stage Feature Extraction (MNIST) 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.

Source

Original publication

Record last updated 11 February 2026

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.