Skip to content
/ Goal Public

Creating a community for turning Valiant's and Rényi's vision into an AI model

Notifications You must be signed in to change notification settings

schizyfos/Goal

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 

Repository files navigation

Goal

Creating a community for the development of AI learning model, outlined in the references bellow.

References

[1] Probably approximately correct (PAC), by Leslie Valiant, https://people.mpi-inf.mpg.de/~mehlhorn/SeminarEvolvability/ValiantLearnable.pdf "The discovery of the notion of computability constituted a new approach to discovering truths about the world. The logician Kurt Gödel generously acknowledged that computability theory has for the first time succeeded in giving an absolute definition of an interesting epistemological notion, i. e., one not depending on the formalism chosen. What can be computed does not change as one varies the details of the model."

[2] A Socratic dialogue on mathematics, by Alfréd Rényi, https://www.cambridge.org/core/services/aop-cambridge-core/content/view/13265932B04BE25A51D3944209C3AA08/S0008439500052346a.pdf/socratic_dialogue_on_mathematics.pdf "Why do you think, Socrates, that this method of thinking and arguing can be used only for the study of numbers and geometric forms? Why do you not try to convince your fellow-citizens to apply the same high logical standards in every other field, for instance in philosophy and politics; in discussing the problems of everyday private and public life?" From that time, this is what I have always tried to do".

[3] PAC learning and Deep nets https://elmos.scripts.mit.edu/mathofdeeplearning/2017/05/08/mathematics-of-deep-learning-lecture-4/

[4] PAC learning in the presence of evasion adversaries https://www.princeton.edu/~pmittal/publications/pac-neurips18.pdf

[5] On the universality of deep learning https://proceedings.neurips.cc//paper/2020/file/e7e8f8e5982b3298c8addedf6811d500-Paper.pdf

[6] PAC confidence sets for deep neural networks via calibrated prediction https://openreview.net/pdf?id=BJxVI04YvB

[7] Learning algorithms from circuit lower bounds https://arxiv.org/abs/2012.14095

[8] Learning algorithms versus automatability of Frege systems https://arxiv.org/abs/2111.10626

Bratislava circle

An international creative circle, (inspired by Vienna circle from the beginning of the last century) of theorists and practitioners to achieve this top noble and interesting goal.

Blog part 1 is here

https://schizyfos.wordpress.com/2021/12/25/251221-conceptual-engineering/

Blog part 2 is here

https://schizyfos.wordpress.com/2021/12/26/261221-conceptual-engineering-2/

Blog part 3 is here

https://schizyfos.wordpress.com/2021/12/28/281221-conceptual-engineering-3/

Blog part 4 is here

https://schizyfos.wordpress.com/2021/12/30/301221-conceptual-engineering-4/

About

Creating a community for turning Valiant's and Rényi's vision into an AI model

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published