Matan Shtepel

on the atlas · 8

• 297.pdf1 savers
• Data availability sampling and danksharding: An overview and a proposal for improvements - a16z crypto1 savers
• In Defense Of Knowing Who You're Dating1 savers
• Every Bay Area House Party - by Scott Alexander12 savers
• The Cryptographer Who Ensures We Can Trust Our Computers | Quanta Magazine1 savers
• deeplearningbook.org/contents/ml.html2 savers
• 275A, Notes 0: Foundations of probability theory | What's new1 savers
• Curius / Onboarding2704 savers

highlights · 11

• he’s read newspapers, biographies, trade press. He went over to his grandfather who was a grocer and he read the progressive

  Curius / Onboarding

• (i) they will only be stored for a limited time, for example, 30-60 days, and (ii) the data will not be directly accessible to smart contracts despite being part of the transaction data. Instead, only a short commitment to the blob data, called DATAHASH

  Data availability sampling and danksharding: An overview and a proposal for improvements - a16z crypto

• Looking for someone who will help me solve the gender imbalance and fertility crisis in Bedroomrundi go on a few dates with me and see what develops (no pun intended).

  In Defense Of Knowing Who You're Dating

• This is inpart because the bounds are often quite loose and in part because it can be quitedifficult to determine the capacity of deep learning algorithms.

  deeplearningbook.org/contents/ml.html

• The VC dimension measures the capacity of a binary classifier. TheVC dimension is defined as being the largest possible value ofmfor which thereexists a training set ofmdifferentxpoints that the classifier can label arbitrarily

  deeplearningbook.org/contents/ml.html

• We can control whether a model is more likely to overfit or underfit by alteringitscapacity. Informally, a model’s capacity is its ability to fit a wide variety offunctions.

  deeplearningbook.org/contents/ml.html

• 1. Make the training error small.2. Make the gap between training and test error small.

  deeplearningbook.org/contents/ml.html

• What separates machine learning from optimization isthat we want thegeneralization error, also called thetest error, to be low aswell.

  deeplearningbook.org/contents/ml.html

• There’s nothing better than being a student.

  The Cryptographer Who Ensures We Can Trust Our Computers | Quanta Magazine

• Machine learning isessentially a form of applied statistics with increased emphasis on the use ofcomputers to statistically estimate complicated functions and a decreased emphasison proving confidence intervals around these functions;

  deeplearningbook.org/contents/ml.html

• Starting [at a young age] he’s read everything that he could find about business.

  Curius / Onboarding