Hacker News Reader

baylearnon Notepad++'s GitHub: 'Free Uyghur' release sparks spam tsunami by pro-Chinese

Glad someone wrote a story to summarize the attack!

Here's the thread about the actual attack on Notepad++'s GitHub issues yesterday by pro-China users:

https://news.ycombinator.com/item?id=21395251

baylearnon Protester shot in chest by live police round during Hong Kong protests

I agree with you. I think the (relatively) success of the movement so far is due to the sheer creative energy of the protester population.

That said I am fine with measured civil disobedience and symbolic destruction of government / quasi-government / tax-payer funded infrastructure when appropriate.

baylearnon Hong Kong Is Winning the Global Public-Opinion War with Beijing

I guess they did.

For posting this article, I'm sorry if it wasn't appropriate for HN for anything, but I think the story around:

"The city’s protest movement has unofficial representatives, crowdfunded advertising, viral videos, and much else that has caught China off guard."

might be interesting to the HN audience.

baylearnon Uyghur woman forced to marry Han Chinese man in exchange for family's release

Related article “China coerces Uighur women into unwanted marriages” with more context:

https://share.america.gov/china-coerces-uighur-women-into-un...

baylearnon Footage of hundreds of blindfolded, shackled Uighur prisoners in Xinjiang, China

A twitter thread from ASPI verifying the authenticity of the YouTube video: https://twitter.com/nrg8000/status/1175353408749891584

baylearnon China is harvesting organs from detainees, tribunal concludes

Oops thank you!

baylearnon [dead]

If an nytimes article (or any article) is not interesting to the HN community, wouldn't it simply not get upvoted and left to decay and rot? Why is a ban needed?

I feel a ban will only serve a minority audience, not the majority of HN who can choose with upvotes.

baylearnon A Conditional Transformer Language Model for Controllable Generation

Paper:

CTRL: A Conditional Transformer Language Mode for Controllable Generation (Largest Publicly Available Language Model)

https://einstein.ai/presentations/ctrl.pdf

Link to model release:

https://www.github.com/salesforce/ctrl

baylearnon DeepPrivacy: A Generative Adversarial Network for Face Anonymization

link to the arxiv paper: https://arxiv.org/abs/1909.04538

baylearnon Hong Kong can be a gateway to liberal values for China

I'd argue that had also things have been going in that direction (to some extent), before Xi Jinping assumed power in late 2012.

baylearnon [dead]

Thanks for your concern, model plumber.

See: https://github.com/hongkonggong/beyond-lennon-walls/blob/mas...

---

Edit: My story got flagged and taken off Hacker News.

baylearnon Xinjiang University President Tashpolat Tiyip Sentenced to Death in Secret Trial

Seriously? (* citations needed.)

baylearnon Xinjiang University President Tashpolat Tiyip Sentenced to Death in Secret Trial

According to Amnesty International, after 2 year suspended death sentence, he could be executed in September of this year: https://www.amnesty.org/download/Documents/ASA1710062019ENGL...

baylearnon Hong Kong Protests: The most striking illustrations from the movement so far

Another thread with more visual art related to the movement: https://twitter.com/hardmaru/status/1163340286983979009

baylearnon Exploring Weight Agnostic Neural Networks

Previous discussion (about the actual research article at https://weightagnostic.github.io/ rather than the blog post):

https://news.ycombinator.com/item?id=20160693

baylearnon Gen: Probabilistic programming system with programmable inference

How is this compared to Turing.jl [1], and Stan[2] /Pyro [3], Edward [4], TensorFlow probability [5]?

Refs:

[1] http://turing.ml

[2] https://mc-stan.org

[3] https://pyro.ai

[4] http://edwardlib.org

[5] https://www.tensorflow.org/probability/overview

baylearnon Robocrop: world's first raspberry-picking robot set to work

I'll believe it when I see a video of the “final version”.

Until then it's just like any other flaky and slow robot grasper demos out there trained using various flaky deep reinforcement learning algorithms.

baylearnon DeepMind readies first commercial product

April Fools!

baylearnon It’s probably time to learn Chinese

Why is Hong Kong (predominantly Cantonese speaking) used as the image for an article about the need to learn Mandarin? Most people in HK do not speak Mandarin natively (although English is spoken almost everywhere here).

baylearnon Show HN: Kuzushiji-MNIST

Link to corresponding paper about Deep Learning for Classical Japanese Literature:

https://arxiv.org/abs/1812.01718

baylearnon Jack Ma Confirmed as Chinese Communist Party Member

How is this news? Most people in China who is a somebody is a party member.

baylearnon World Models with a Reservoir Computing Twist

Background information about Reservoir Computing: https://en.wikipedia.org/wiki/Reservoir_computing

World Models paper: https://arxiv.org/pdf/1803.10122.pdf https://worldmodels.github.io

Previous HN Discussion: https://news.ycombinator.com/item?id=16860247 https://news.ycombinator.com/item?id=16860247

baylearnon Show HN: I trained a neural network to write Kanji

How so?

baylearnon Grid-Cell Like Deep Layers in Navigation AI

Yes, refer to the openreview link to my earlier comment on this thread.

baylearnon Grid-Cell Like Deep Layers in Navigation AI

This DeepMind work has been gathering a lot of hype, but it is very similar to this work by Cueva and Wei:

Published on Arxiv on March 31: https://arxiv.org/abs/1803.07770

Published at ICLR 2018, a well-known machine learning conference: https://openreview.net/forum?id=B17JTOe0-

It is rather disheartening that they didn't acknowledge the -- seems like a deliberate oversight.

The Cueva and Wei paper is well written, explains concepts clearly, without the hype, and published in an open access venue.

baylearnon Statement on Nature Machine Intelligence

Nature does have an "Open Access" option in their policy:

"Authors wishing to publish open access retrospectively should contact the individual journal editorial office, and will need to sign a new open access licence to publish form (LTP) and pay the journal's article processing charge."

http://www.nature.com/openresearch/about-open-access/policie...

However, they don't disclose how much money it costs to convert one's work to open access, and whether this option is available for all Nature journals, the main Nature journal, or just the specialized subject journals, such as Nature Machine Intelligence.