ryptophan

Probably yes - and I think my results suggest that optimizing for each would not only mean different endpoints but also different embedding models behind each. One model may measure best for short human-authored queries, while a different model may provide better results for longer LLM-authored query forms.

Author here. I'm a Tech Lead at Georgian (a growth-stage VC).

Short version: I used a few retrieval benchmark datasets, and experimented with a 21 embedding model leaderboard. I held the benchmark corpus and relevance labels fixed, and transformed the queries: paraphrase, terse, verbose. I kept the “information need” the same under each query transformation, but the leaderboard order changed more than I expected.

I reproduced published benchmark scores for benchmark dataset + embedding model pairs, and then used that same setup with transformed queries as the input. I observed the reordering under each rewrite style to be consistent across 4 independently-generated rewrites (gemini-2.5-pro, claude-opus-4.8, gpt-5.1, llama-3.3-70b) rather than showing up in just one.

Code and query score files are here so you can check the tables without a full re-run: https://github.com/georgian-io/query_form_experiment

Let me know your thoughts on this … or in general let me know about any experience with putting an embedding model into a production retrieval system and being surprised by poor performance.

Dermatology is complicated. A labeled image dataset of skin conditions?

Ooo this has been on my mind as well. To add, if I may, off the top of my head:

- How do your taxes work (I would assume not deducted by payroll, etc)

- If you enter USA for any work-related thing, do you need a work visa? (I assume the answer is yes, but maybe it's more nuanced?)

To the general HN community: has anyone seen/heard of US companies hiring up north to try and get a 'discount' on talent, based on the currently weak CAD?

Host frontend on s3, use route53 to connect domain, connect form to AWS lambda function. Should be less than $2 a month.

AWS currently not offering Lightsail/EC2 hosting within Canada, as well. DigitalOcean has been in Toronto for a while. Certainly not an important detail to many, but it was for me.

AWS has been saying they will open in Canada for quite a while... seem to keep pushing the date further and further.

I'll get to work on it. It'll cost you $2/month, though.

How much of Watson's NLP success can be attributed to IBM acquiring AlchemyAPI (http://www.alchemyapi.com/)?

I mean, it's cool and all... but I'm not sure I'd consider the input images to be rough sketches.

To me it's more like a pencil -> ink converter.

Haven't tried Azure yet - will give it a spin eventually. I love Digital Ocean compared to AWS. Sure, AWS has a lot of bells and whistles, but this simplicity of DO is just what I need most times.

curl https://cdn.jsdelivr.net/cash/1.0.0/cash.min.js | wc -c # 7769

curl https://cdn.jsdelivr.net/cash/1.0.0/cash.min.js | gzip | wc -c # 2636

Another issue that seems to be a major hangup in my jurisdiction (which has banned Uber from operating) is the lack of "proper insurance" for Uber drivers/cars.