smougel
Re:Why do we need dithering?
on Friday November 14, 2025 @07:35AM (#45926165)
Just Vibecoded this : https://x.com/smougel/status/1989309185423532120?s=20
Subpixel dithering ! 1 bit per channel which mean that what you see is only 0 or 1 for each channel (R, G, B) By using gaussian blur, the result is perceptually very good ! X compress the image a lot but this truly 1 subpixel ON / OFF
Re:Two things LLM coding agents are still bad at
on Thursday October 09, 2025 @08:14AM (#45526554)
Not related to code... But when I use a LLM to perform a kind of copy/paste, I try to number the lines and ask it to generate a start_index and stop_index to perform the slice operation. Much less hallucinations and very cheap in token generation.
Re:The Einstein AI Model
on Tuesday March 11, 2025 @06:45AM (#43331123)
The current culture about AI & LLMs is that we are "only" memorizing the Web into model parameters and that a LLM is unable to "invent" new paradigms. Maybe we are under estimating what Unsupervised Learning & RL could provide. Re-inforcement learning is about exploring and finding new ways to accomplish a task and I see no limit here (except the computational resources needed).
Re:You Should Probably Pay Attention to Tokenizers
on Thursday October 24, 2024 @09:39AM (#41935365)
There is a paper about : https://arxiv.org/pdf/2405.18719
Re:Show HN: What are content semantic maps?
on Friday January 12, 2024 @03:07AM (#38965398)
Content semantic maps allows to visualize content of a website in a 2d space.
It allows new kind of usage : competitor analysis, conversion visualization, content gap analysis.
Re:Show HN: SEOCopilot – Find keywords easy to rank and that converts
on Wednesday September 06, 2023 @09:23AM (#37404758)
Keywords recommendations are refreshed every night.
You can expect some fluctuation depending of several factors (search queries extracted from Google search console etc.)
XLS Keyword export allows up to 10000 records extraction.
Show HN: SEOCopilot – Find keywords easy to rank and that converts
No magic here, just data science. Around 80% of web content yields no conversions. Our goal: To identify, among millions of search queries, those that lead to conversions (such as purchases, contact form submissions, or app downloads).Technologies involved: Google Search Console, Transformers, Clustering, Semantic Similarity.For more details, visit:...
Re:MSPaint in JavaScript
on Wednesday February 26, 2020 @09:06AM (#22422991)
So true!
Re:MSPaint in JavaScript
on Wednesday February 26, 2020 @09:05AM (#22422986)
Huge work!
That's awesome!
The power of photoshop for Linux. No install. Supra fast!
Wow!
Re:PHP Apprentice – An online, open-source book for learning PHP
on Friday January 11, 2019 @11:19AM (#18884510)
Good job ! Please add something about namespaces.
Re:23andMe Informed Me My Husband and I Are Related
on Friday December 14, 2018 @05:37AM (#18679854)
> With this company you end up $100 (or whatever) short and you surrender your data for it to be sold to ... .... insurance companies and whatnot.
False ! Source : https://www.23andme.com/about/privacy/
We will not provide any person’s data (genetic or non-genetic) to an insurance company or employer.
Re:Fundamental Algorithms of Machine Learning
on Thursday November 22, 2018 @09:26AM (#18510135)
Far more complete : https://www.deeplearningbook.org/
(Ian Goodfellow and Yoshua Bengio and Aaron Courville)
Re:Show HN: An API to extract texts from images and PDF files
on Thursday July 18, 2013 @10:46AM (#6063967)
Any Feedback Welcomed
Show HN: An API to extract texts from images and PDF files
Re:Ask HN: In which cases you would have liked a PDF to be editable?
on Thursday May 02, 2013 @08:29AM (#5643989)
Each time I need to open Photoshop, modify, and print to a PDF File...