Andreas Eriksen

He/him. Dad and husband. I work at My main hobbies are sailing, mountain/road biking, and creating an ultra low power personal computer.

[Andreas Eriksen @andreer]

@bclavie @antoine_chaffin We're ready lol https://t.co/P97WnfGsno

[Julian Killingback @Julian_a42f9a]

Hypencoder is now ready for production using Vespa! Props to @andreer for making it happen! Really great stuff!

[Andreas Eriksen @andreer]

@antoine_chaffin @bclavie loving this new kind of verifiable vagueposting

[Andreas Eriksen @andreer]

@capemox @raphaelsrty Ah, awesome. Are you planning to release embedding models based on these too? I was thinking it would be amazing to have the whole stack fully open and reproducible

[Andreas Eriksen @andreer]

@capemox @raphaelsrty Thanks for publishing the pretrained checkpoints! I'm also pretty resource constrained when playing with this stuff, partly by VRAM but mostly by lack of patience with longer jobs lol

[Andreas Eriksen @andreer]

yeah why force all the data to flow through the lossiest, most expensive layer

[Yohei @yoheinakajima]

slowly we're all realizing that tools should be called from code, not from within the llm api

[Andreas Eriksen @andreer]

@bclavie nah keep the good vibes going

[Dainius Jocas @dainius_jocas]

I’m excited to join the speaker lineup at https://t.co/QxAGy0IKON Live! My session will explore Nuances of Binarized Embeddings-Based Retrieval

If you’re attending, let me know — would love to connect at the event.

[Erik Bernhardsson @bernhardsson]

My opinion on tokenmaxxing is companies shouldn’t mandate/constrain any tools at all and then evaluate software developers by output / (salary + token use)

[Andreas Eriksen @andreer]

@bclavie @din0s_ lol

[vespa.ai @vespaengine]

Vespa isn't only powering American AI, now it's also @MistralAI's retrieval solution

[Thomas Thoresen @thomas_thoresen]

bm25 is nice and all, but you won't believe how easy it is to improve upon it with and how much more you can squeeze from lexical features in @vespaengine

[Andreas Eriksen @andreer]

Bound to happen, very happy to see it!

[Omar Khattab @lateinteraction]

Claude Code is finally an RLM (oct 2025), congrats to Anthropic :-)

[Andreas Eriksen @andreer]

@tomaarsen @LightOnIO I had started with bge-reranker-v2-m3 so this is an awesome free upgrade! Thanks for publishing

[Andreas Eriksen @andreer]

@tomaarsen @LightOnIO Had already noticed the scores had good room for improvement

[Andreas Eriksen @andreer]

@tomaarsen @LightOnIO Awesome - I just killed my ongoing rescoring job started 3 hours ago, and started the download instead. 😁

[Andreas Eriksen @andreer]

Yeah this was great. Also feel a deep connection with wanting to do stuff but not feeling qualified lol

[John Carmack @ID_AA_Carmack]

I loved the Project Hail Mary movie (and the book).

Science, technology, competence, and openness – this is my culture.

[Andreas Eriksen @andreer]

PSA: btop is amazingly nice. Among the first things I install on any linux box I use

[Kyle Boddy @drivelinekyle]

Long live btop!

[Andreas Eriksen @andreer]

@tomaarsen Well, consider that part resolved 😉We "supported" it since long before the paper dropped, this could likely run just as well on years old Vespa version. Ofc would love to optimize for it specifically but that is likely gated on usage

[Andreas Eriksen @andreer]

Just added a sample app for how to search with hypencoder models on Vespa. A large meta-model that generates a small query-specific model that scores your docs - it feels like pure science fiction, but of course we can do it: https://t.co/3TAE6dnTvQ

[Andreas Eriksen @andreer]

@tomaarsen Of course that is purely theoretical unless we can show benefit from that thru training, more compelling results would be needed to raise interest