Vicki
@vickiboykis.com
Founding list[float] engineer. Recsys. Personalization. Infra. Systems. Normcore code. Nutella. Vectors. Words. Vibes. Bad puns (soon).
https://vickiboykis.com/what_are_embeddings/
https://vickiboykis.com/what_are_embeddings/
When you’re on your fourth merged PR for the week
September 24, 2025 at 12:26 PM
When you’re on your fourth merged PR for the week
philosophical question, what would you consider the dim size of a matryoshka embedding?
August 9, 2025 at 2:48 PM
philosophical question, what would you consider the dim size of a matryoshka embedding?
thinking i need to update this part of the embeddings paper, this was the case when i wrote it but likely no longer true - is everyone using 768 (min) now?
August 9, 2025 at 2:28 PM
thinking i need to update this part of the embeddings paper, this was the case when i wrote it but likely no longer true - is everyone using 768 (min) now?
Trick question, it will never scale perfectly. This is how they tricked us into AWS lambda
mirrors.edge.kernel.org/pub/linux/ke...
mirrors.edge.kernel.org/pub/linux/ke...
August 4, 2025 at 10:35 PM
Trick question, it will never scale perfectly. This is how they tricked us into AWS lambda
mirrors.edge.kernel.org/pub/linux/ke...
mirrors.edge.kernel.org/pub/linux/ke...
This was the phishing email I got, which had links to my actual Twitter settings. If you get something similar, don't click!
July 23, 2025 at 2:20 AM
This was the phishing email I got, which had links to my actual Twitter settings. If you get something similar, don't click!
July 7, 2025 at 12:00 PM
![tzury 5 hours ago | prev | next [–]
I’ve found that starting with a plain old filesystem often outperforms fancy services - just as the Unix philosophy (“everything is a file” [1]) has preached for decades [2].](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:b65l7jiwk75kx3woz5ewhyib/bafkreie64flthmr63zammeoqnixwblbqtjrjqb7wo46jbksoniym2knn2m@jpeg)
May 5, 2025 at 10:50 PM
March 20, 2025 at 11:24 PM
can’t even let the old neurons rot for 20 min on insta
March 19, 2025 at 2:41 AM
can’t even let the old neurons rot for 20 min on insta
Ah yeah. In that case I'd probably structure it mostly something like this (depending on the language,what my users expect) docs.astral.sh/uv/concepts/...
There's a top-level project (not sure I'd always use "packages", but bird-feeder could be the app and seeds the API/SDK)
There's a top-level project (not sure I'd always use "packages", but bird-feeder could be the app and seeds the API/SDK)
February 27, 2025 at 2:30 PM
Ah yeah. In that case I'd probably structure it mostly something like this (depending on the language,what my users expect) docs.astral.sh/uv/concepts/...
There's a top-level project (not sure I'd always use "packages", but bird-feeder could be the app and seeds the API/SDK)
There's a top-level project (not sure I'd always use "packages", but bird-feeder could be the app and seeds the API/SDK)
is this for the search UI or the results? I find it to be pretty good, one of the only good interfaces for industrial RAG...so far. The cited sources really are what do it for me.
February 27, 2025 at 2:26 PM
is this for the search UI or the results? I find it to be pretty good, one of the only good interfaces for industrial RAG...so far. The cited sources really are what do it for me.
Just rereading some recsys papers from 2022, could be blanket-applied to LLM evaluation today
ojs.aaai.org/aimagazine/i...
ojs.aaai.org/aimagazine/i...
February 26, 2025 at 1:16 PM
Just rereading some recsys papers from 2022, could be blanket-applied to LLM evaluation today
ojs.aaai.org/aimagazine/i...
ojs.aaai.org/aimagazine/i...
damn this is super slick
February 7, 2025 at 1:42 PM
damn this is super slick
How do we think they picked these?
January 21, 2025 at 1:55 AM
How do we think they picked these?
Non è possibile fuori dall'italia :/
January 13, 2025 at 11:38 AM
Non è possibile fuori dall'italia :/
Found this absolute gem of a texbook on NLP recently, with context from the last few years
web.stanford.edu/~jurafsky/sl...
web.stanford.edu/~jurafsky/sl...
January 11, 2025 at 3:37 PM
Found this absolute gem of a texbook on NLP recently, with context from the last few years
web.stanford.edu/~jurafsky/sl...
web.stanford.edu/~jurafsky/sl...
Programming for about ten years
January 7, 2025 at 3:33 AM
Programming for about ten years
Getting cleaner and cleaner, unfortunately still doesn't do present tense well :\
January 6, 2025 at 11:55 PM
Getting cleaner and cleaner, unfortunately still doesn't do present tense well :\
Vibe check is ok on smaller diffs, falls apart on larger ones but the goal is to also make this into a runnable more scientific test suite for this kind of stuff.
January 6, 2025 at 5:30 PM
Vibe check is ok on smaller diffs, falls apart on larger ones but the goal is to also make this into a runnable more scientific test suite for this kind of stuff.
Making some progress here, small pretty runnable that generates potential commit messages for you using various models via the ollama APi
January 6, 2025 at 12:41 AM
Making some progress here, small pretty runnable that generates potential commit messages for you using various models via the ollama APi
What I'm wondering is, how do I know that the places suggested by Gemini/ChatGPT are real places? With Perplexity it's easy because I have the links.
January 4, 2025 at 2:32 AM
What I'm wondering is, how do I know that the places suggested by Gemini/ChatGPT are real places? With Perplexity it's easy because I have the links.
Does it depend on the model? Haven't been able to trigger it
January 4, 2025 at 1:40 AM
Does it depend on the model? Haven't been able to trigger it