HN user

evc123

308 karma
Posts35
Comments26
View on HN
news.ycombinator.com 3y ago

“Broken Neural Scaling Laws”; New Scaling Law; SotA Extrapolation of All Things

evc123
2pts0
medium.com 8y ago

OpenAI Retro Contest Report

evc123
1pts0
www.reddit.com 8y ago

What do we already know about Generalization? What should we ask next about it?

evc123
1pts0
www.graphcore.ai 8y ago

Revisiting Small Batch Training for Deep Neural Networks

evc123
2pts0
arxiv.org 8y ago

Revisiting Small Batch Training for Deep NNs; batch_size less than 64 is best

evc123
1pts0
www.biorxiv.org 8y ago

Prefrontal cortex as a meta-reinforcement learning system [DeepMind]

evc123
3pts0
petewarden.com 8y ago

The Machine Learning Reproducibility Crisis

evc123
2pts0
www.youtube.com 8y ago

Pan-Asian Deep Learning Conference; Kuala Lumpur, Malaysia; 2028

evc123
1pts0
github.com 8y ago

PyTorch v0.2.0 released

evc123
150pts14
www.youtube.com 8y ago

Graphcore: How to build a processor for machine intelligence

evc123
1pts0
blog.aqnichol.com 9y ago

Why I’m Remaking OpenAI Universe

evc123
224pts58
www.youtube.com 9y ago

Facebook's brain-computer interface presentation

evc123
1pts0
devblogs.nvidia.com 9y ago

Recursive Neural Networks with PyTorch

evc123
2pts0
scaledml.org 9y ago

Recent Progress in Generative Modeling – Ilya Sutskever OpenAI [pdf]

evc123
2pts0
arxiv.org 10y ago

[1606.03813] Towards an integration of deep learning and neuroscience

evc123
1pts0
www.technologyreview.com 10y ago

Basic income a sellout of the american dream; critiques YC basic income research

evc123
2pts2
arxiv.org 10y ago

Dialog-Based Language Learning [arXiv:1604.06045] (by Jason Weston @ fb ai)

evc123
5pts0
arxiv.org 10y ago

A Parallel-Hierarchical Model for Machine Comprehension on Sparse Data

evc123
3pts1
blogs.microsoft.com 10y ago

Microsoft Is Using Minecraft for AI Research

evc123
2pts0
medium.com 10y ago

CreativeAI: On the Democratisation and Escalation of Creativity – Chapter 01

evc123
4pts0
arxiv.org 10y ago

Dynamic Memory Networks for Visual and Textual Question Answering

evc123
122pts15
www.youtube.com 10y ago

Artificial vs. Biological Systems

evc123
1pts0
cs231n.stanford.edu 10y ago

Videos and Slides for Stanford CS231n: Convolutional Neural Networks for Vision

evc123
17pts0
www.itshared.org 10y ago

Data science interview questions

evc123
4pts0
drive.google.com 10y ago

Obstacles on the Path to AI

evc123
129pts29
www.youtube.com 10y ago

Eli the Computer Guy Says YC Needs to Teach Better Business Communication Skills

evc123
2pts0
arxiv.org 11y ago

Ask Me Anything: Dynamic Memory Networks for Natural Language Processing

evc123
12pts2
www.youtube.com 11y ago

Bill Gross: The single biggest reason why startups succeed

evc123
3pts0
arxiv.org 11y ago

End-to-End Training of Deep Visuomotor Policies [pdf]

evc123
5pts0
www.youtube.com 11y ago

Deepmind office interview

evc123
2pts0
OpenAI Charter 8 years ago

What about benefitting non-human animals? Hopefully the benefits are distributed to all creatures and not just humanity.

Evolution has been running ~10^19 experiments in parallel for billions of years: http://reducing-suffering.org/how-many-wild-animals-are-ther...

Evolution is a slow algorithm, but it had access to an absurd amount of compute (all neuronal organic matter on Earth) and environment simulation (all of physical reality on Earth) when discovering us; so the discovery of the algorithms/architectures/principles in our heads shouldn't be viewed as trivial.

Special projects 10 years ago

@Darmani Can you think of any ways traditional program synthesis techniques could be combined with machine learning to perform #2? Assume your system has access to a large amount of practice problems/solutions to train with.

Was his cerebral hemorrhage a sudden/unpredictable death? If not, I'm surprised he didn't first relocate to cooperative hospice care in Scottsdale near Alcor to cryopreserve himself before damage became (possibly) irreversible.

I think Facebook has staying power because they're the first social network to have made serious AI/ML investments. They'll be able to project user's likes back at them better than anyone else.