Markus Zimmermann (@zimmskal) 's Twitter Profile
Markus Zimmermann

@zimmskal

Benchmarking LLMs to check how well they write quality code. Support me using the profile link 👇

ID: 218911998

linkhttps://buy.stripe.com/5kA3g962hfP0dGMeUX calendar_today23-11-2010 13:59:30

3,3K Tweet

2,2K Followers

884 Following

Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

Going on vacation without a laptop for the first time since... 10 years?!? But the most exciting thing is by far seeing my oldest child be super excited and packing all the things she would like to take with. Not that we have room for literally every toy but still 💘

Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

I see that my request for not releasing a major model during the easter holidays was fully ignored 😿 Well, here we go 🏇 If somebody knows how i get i free and not rate limited token for benchmarking OpenAI's o3, please let me know.

I see that my request for not releasing a major model during the easter holidays was fully ignored 😿

Well, here we go 🏇

If somebody knows how i get i free and not rate limited token for benchmarking <a href="/OpenAI/">OpenAI</a>'s o3, please let me know.
Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

Still believe that this is the development process to go, even in a coding agent world: symflower.com/en/company/blo… But... especially because of what i have heard lately about development processes...

Still believe that this is the development process to go, even in a coding agent world: symflower.com/en/company/blo…

But... especially because of what i have heard lately about development processes...
Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

I just got demoed a new amazing model and was asked about my favorite question i usually prompt. I used to have one that is not up-to-date-data or coding related when the first reasoning models came out: `Actually create a proof for the P versus NP problem. Make a plan on how to

Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

Our kid number 2 is a super simple finite state machine: - drink - eat - drink - play - poop - sleep - repeat Deviate from that master plan and you will get screamed at 🫡

Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

"You know, I think we'll get to full self-driving next year. As a generalized solution, I think" Sorry Elon Musk but I think you must come up with a new repeatable future prediction quote like right now!

Markus Zimmermann (@zimmskal) 's Twitter Profile Photo

Feature request willhaben: fixe Treffen mit Ort Datum und Zeit die verpflichtend sind für beide Seiten. Wenn du Person nicht innerhalb von $agreeable-delay da ist. Dann kassiert die andere Person $amount. E.g. 5 Euro. Ich wäre schon Millionär damit.