pooyak
Re:Don't use cosine similarity carelessly
on Tuesday January 14, 2025 @06:35PM (#42705455)
Very interesting article. Is there any model that can generate embeddings given a system prompt? This can be useful not only for similarity searching but also for clustering use cases without having to do too much custom work. Essentially, a zero shot embedding model.
Here is one version of 4D that our brain understands and can easily map from 4D back to 3D, and that's when the 4th dimension is time (or motion): http://www.dyscario.com/arts-and-culture/sculptures-in-motio...
Re:Hyperloop Alpha [pdf]
on Monday August 12, 2013 @07:27PM (#6202927)
I converted the PDF to a rough markdown:
https://github.com/pooyak/hyperloop/blob/master/document/hyp...
Needs more clean up, but still easier to work on.
Re:The People's Bailout
on Thursday November 08, 2012 @10:33PM (#4761126)
How does buying debt from creditors work? Would something like this actually cause buying debt more expensive and thus benefit creditors instead?
Re:Diffusion Curves: A Vector Representation for Smooth-Shaded Images
on Wednesday October 31, 2012 @01:46PM (#4724132)
Did you try to compress the resulting SVG files? Gzipping them will usually make them 3-4x smaller.
Re:NameCheap Overtakes Go Daddy In Google
on Friday April 20, 2012 @12:54PM (#3868697)
Just a few days ago, I was about to move to namecheap but then I find out their interface was a little too simple. It was missing more advanced features that even godaddy has like import/export zones. Looking for other suggestions too.
Re:Prince of Persia source code released (Apple II assembly)
on Tuesday April 17, 2012 @04:14AM (#3851696)
wait! no "megahit" in source?
this can't be real...
Re:Common Crawl
on Sunday March 11, 2012 @09:04PM (#3691891)
This blog post:
http://matpalm.com/blog/2012/01/01/common_crawl_collocations...
mentions that commoncrawl's data is last updated September 2010. Does anyone know if that's still the case?
Re:CommonCrawl: an open repository of web crawl data that is universal...
on Tuesday December 13, 2011 @12:03AM (#3346377)
thread on HN from when common crawl was announced, interesting info there:
http://news.ycombinator.com/item?id=3209690
Re:Free 5 Billion Page Web Index Now Available from Common Crawl Found...
on Tuesday November 08, 2011 @01:24AM (#3209732)
One interesting discussion from here:
http://www.commoncrawl.org/common-crawl-enters-a-new-phase/
It says the cost of running a hadoop job to scan all 5billon documents is in the order of $100.
Does any one know how does this compare to let say Yahoo BOSS? Is it even comparable?