Janelle Shane

The neural network doesn’t understand ingredients

By Janelle Shane On March 02, 2016 - 1 min read

As the neural network learns to generate recipes [http://lewisandquark.tumblr.com/post/140219420017/the-silicon-gourmet-training-a-neural-network-to] , it’s very clear from the ingredients it produces that it has no clue what these words actually mean, or whether the substances involved are powders, liquids, solids,

The neural network doesn’t understand pepper.

By Janelle Shane On March 01, 2016 - 1 min read

As the neural network tries to learn to generate recipes by looking at a 30MB cookbook [http://lewisandquark.tumblr.com/post/140219420017/the-silicon-gourmet-training-a-neural-network-to] , there are certain ingredients it has to learn to deal with. One of the most confusing for the network is

What is teaspoons?

By Janelle Shane On March 01, 2016 - 1 min read

I’m training a neural network to generate cookbook recipes [http://lewisandquark.tumblr.com/post/140219420017/the-silicon-gourmet-training-a-neural-network-to] . It looks at a bunch of recipes and has to figure out completely from scratch - no knowledge of what English words even are - how to start

The Silicon Gourmet: training a neural network to generate cooking recipes

By Janelle Shane On February 29, 2016 - 3 min read

Neural networks are computer learning algorithms that mimic the interconnected neurons of a living brain, managing astonishing feats of image classification [http://googleresearch.blogspot.com/2015/06/inceptionism-going-deeper-into-neural.html] , speech recognition [https://www.youtube.com/watch?v=yxxRAHVtafI], or music generation [https://highnoongmt.wordpress.com/2015/

Friday November 13, 2015

By Janelle Shane On November 13, 2015 - 1 min read

Looks like a slanted surface of Martian desert cliffs, but this is actually an extreme closeup of a human hair. Nanoresearchers who try to describe their research often up comparing their devices to the size of a human hair - for good reason. At about 100 micrometers thick, a human hair

(Untitled)

By Janelle Shane On November 12, 2015 - 1 min read

Looks like a slanted surface of Martian desert cliffs, but this is actually an extreme closeup of a human hair. Nanoresearchers who try to describe their research often up comparing their devices to the size of a human hair - for good reason. At about 100 micrometers thick, a human hair

Sunday September 06, 2015

By Janelle Shane On September 06, 2015 - 1 min read

ucresearch [http://ucresearch.tumblr.com/post/128410787640]: > These are not the latest Muppets characters, but microscopic lasers that have been digitally enhanced with some googly-eyes. The nanolasers are built in the engineering labs at UC San Diego, and according to Janelle Shane, they have the potential to increase

(Untitled)

By Janelle Shane On September 06, 2015 - 1 min read

ucresearch [http://ucresearch.tumblr.com/post/128410787640]: > These are not the latest Muppets characters, but microscopic lasers that have been digitally enhanced with some googly-eyes. > The nanolasers are built in the engineering labs at UC San Diego, and according to Janelle Shane, they have the potential to

Saturday August 29, 2015

By Janelle Shane On August 29, 2015 - 1 min read

These are microscopic balls of tin, imaged under scanning electron microscope (SEM) at 2500x (about 25x more magnification than the strongest optical microscope can manage). These very cool-looking tin samples are good for looking at when you’re trying to calibrate the SEM to produce the best images possible

(Untitled)

By Janelle Shane On August 29, 2015 - 1 min read

These are microscopic balls of tin, imaged under scanning electron microscope (SEM) at 2500x (about 25x more magnification than the strongest optical microscope can manage). These very cool-looking tin samples are good for looking at when you’re trying to calibrate the SEM to produce the best images possible

The neural network doesn’t understand ingredients

The neural network doesn’t understand pepper.

What is teaspoons?

The Silicon Gourmet: training a neural network to generate cooking recipes

(Untitled)

(Untitled)

(Untitled)

Writing

Subscribe

Recent Posts

It's 11:00 pm. Do you know where your AI agent is?

Get working on your April Fools Eiffel Tower

Bonus: More April Fools pranks from Eiffel Tower Llama

When a chatbot runs your store

Bonus: Incorrect Christmas Carols

Tiny neural net Halloween costumes are the best

Follow

It's 11:00 pm. Do you know where your AI agent is?

Get working on your April Fools Eiffel Tower

Bonus: More April Fools pranks from Eiffel Tower Llama