DeepLearningThesis

One metric to ruler them all.

[DeepLearningThesis @deep_thesis]

My Crazy New Idea: It’s Like Uber For Lamps. https://t.co/uvzRyEjgDV

[DeepLearningThesis @deep_thesis]

Downloading Felis catus, 137 MB #ImageNet #ComputerVision

[DeepLearningThesis @deep_thesis]

The guys at @Stanford and @Princeton should really check the performance of the #ImageNet website, incredibly slow response time... #DeepLearning #ComputerVision

[DeepLearningThesis @deep_thesis]

How does one know that one's model has converged? And to which minima? Which effects does this have? Is it possible to compare minima between models? #DeepLearning #deepthoughts

[DeepLearningThesis @deep_thesis]

We're taking part in the #GAIA conference today, great to see an active #AI community in Gothenburg #gbgtech

[DeepLearningThesis @deep_thesis]

great blog post by @parasdahal explaining clearly how Softmax and Cross Entropy Loss actually works https://t.co/f6ceskWBsB

[DeepLearningThesis @deep_thesis]

Great pod by @twimlai with @goodfellow_ian and Sandy Huang - you guys are so pedagogic and professional, we need more people like you! Recognised the dog as fish example from the Influence Functions paper btw, great paper! https://t.co/ewf1sFuAI0 https://t.co/V1MXh4IgN0

[The TWIML AI Podcast @twimlai]

Today we're joined by @goodfellow_ian of @GoogleBrain @googleresearch and Sandy Huang, PhD Student at @UCBerkeley, to discuss their paper "Adversarial Attacks on Neural Network Policies." Follow the link to check it out! https://t.co/QoVbDwvygT

[DeepLearningThesis @deep_thesis]

Us debugging legacy influence function code. Hard at work or hardly working? #InfluenceFunctions #TensorFlow #DeepLearning

[DeepLearningThesis @deep_thesis]

@EliotAndres @ProjectJupyter First example works, thanks!!

[DeepLearningThesis @deep_thesis]

@EliotAndres do you know any easy way to change keyboard shortcuts in @ProjectJupyter? cannot use comment shortcut on non US keyboard, and this feels a bit inconvenient https://t.co/sKmS8Abbiz

[DeepLearningThesis @deep_thesis]

@EliotAndres @ProjectJupyter I seem to be able to edit the Command Mode shortcuts, but not the Edit Mode shortcuts

[DeepLearningThesis @deep_thesis]

A bit creepy how much traction something like this can have in such a short time if it comes from someone as influential as @elonmusk. Would everyone donate to end hunger in Somalia if @elonmusk just said the word?

[Martin Ford @MFordFuture]

Flying taxi drones from @kittyhawkcorp are now being tested in New Zealand
#RiseoftheRobots
@andrewrsorkin https://t.co/souJP2TGxx

[DeepLearningThesis @deep_thesis]

great blog post by @EliotAndres introducing some very useful @ProjectJupyter extensions! https://t.co/HJGyKTKrxn

[Eliot Andres @EliotAndres]

🤓 New blog post: My 5 favorite Jupyter Notebook Extensions: https://t.co/Qi2oEdlKHW With pieces of advice (tweets) from @radekosmulski and @randal_olson

[DeepLearningThesis @deep_thesis]

Great @ProjectJupyter #magic for #debugging https://t.co/rLYK6PCUfU

[Radek Osmulski @radekosmulski]

Favorite recent jupyter notebook discovery - the %debug magic:

1. Get an exception.
2. Insert a new cell, type %debug and run it.

An interactive debugger will open bringing you to where the exception occurred and allowing you to look around!

[DeepLearningThesis @deep_thesis]

Great pod by @twimlai talking to Ksenia Konyushkova about her work in #ActiveLearning etc. Wonder if similar approach could be taken to similarity in trained models, or if influence functions could be applied to active learning #LearningActiveLearning #DeepLearning https://t.co/dnSmpSeurU

[The TWIML AI Podcast @twimlai]

Live!! Today we're joined by Ksenia Konyushkova, PhD Student in the CVLab at @EPFL_en, with whom we discuss her interesting research into ways we might apply machine learning to ease the challenge of creating labeled datasets for machine learning. https://t.co/UBOpUfqRIg

[DeepLearningThesis @deep_thesis]

Yesterday we took part in #hashcode at the #NewMinds hub in #gothenburg! We didn't win, but managed to submit a solution, which made us happy enough! #HappyHacking! #DataScience #hashcode18

[DeepLearningThesis @deep_thesis]

Interesting thread about what #AI actually can do https://t.co/cg3uudnEWv

[Rodney Brooks @rodneyabrooks]

AI learning performance is not yet even in the same continent as biological learning. Good to see people exploring the deep science of learning still. Readable summary: https://t.co/eUa3jysqYc and source material: https://t.co/vTQOg6J74Z

[DeepLearningThesis @deep_thesis]

Nice inspiration from @IBMWatson @xprize regarding #AI for global good. Their approach is much "track everything that can be tracked, make everything IoT". Wonder how to make it more concrete.. https://t.co/oFAANYJ3IW

[Anders Frick @frick]

Bakom scenen men framför kameran idag hos @EFNTV - Daniel @Langkilde från @Annotell för ett snack om bl.a. #artificiellintelligens och #maskininlärning. #gbgtech #gbgstartup #startupgbg