kirtivr

5 ms·

31.
On-call isn't supposed to be this hard(fixbugs.ai)
4 points by kirtivr 5mo ago|1 comments
32.
MiniMax M3 Benchmarks One Pager(filecdn.minimax.chat)
3 points by kirtivr 5mo ago|1 comments
33.
by kirtivr 5mo ago
Some super relevant benchmarks like Humanity's Last Exam, Long context reasoning (MRCR 128K-256K) are not included. Overall this seems to be a strong agent-oriented model. What are the benchmarks that most closely track model coding pe
34.
by kirtivr 5mo ago
They got 1K tok/s with Deepseek v4 Pro. That's kinda cool..
35.
by kirtivr 5mo ago
I can think of real time video, shader generation, real time worldbuilding type problems could require such a high token throughput. For instant code generatio, 400-500 tok/s should be sufficient, though most frontier models give us cl
36.
by kirtivr 5mo ago
OpenRouter and the provider sign a contract clearly specifying how input data is to be handled. It's the same way we trust OpenAI to not train on our data if we've opted out although there is no control on whether they can retain
37.
by kirtivr 5mo ago
How did this article (1) Get published in Science (2) Get 200+ upvotes on HN, with no explanation of the supposed science behind writing? What's the null hypothesis? Is it possible to disprove your assertion?
38.
by kirtivr 5mo ago
You don't need to know who developed the LLM - whether it was Google or OpenAI. What you need to know is who is the provider for the LLM, and whether their endpoints are zero data retention enabled and opted out of training. OpenRouter
39.
by kirtivr 5mo ago
RE: Competing clones of your prodcut If you could go back in time, would you still make the keyboard firmware fully opensource? I am sure having the product open source would have helped keep the community going during the early stages. I&#
40.
by kirtivr 5mo ago
TL;dr Make a first draft that is bad, and improve it from there.
41.
Opus 4.8 System Card [pdf](cdn.sanity.io)
6 points by kirtivr 5mo ago|1 comments
42.
by kirtivr 5mo ago
Training for Mythos finished in February, 2026 while training for Opus 4.7 finished around that same time. If I understand correctly, Opus 4.7 was launched as nerfed Mythos with some improvements from 4.6. Anthropic launches major bumps (li
43.
by kirtivr 5mo ago
I was a regular Claude Code user but Codex eventually won me over due to a few factors: 1. Less interaction required over long horizon tasks. 2. You actually get the amount of tokens they advertize. It's been an open secret on r/C
44.
by kirtivr 5mo ago
For enterprise customers If your organization uses Gemini CLI or our IDE extensions via a Gemini Code Assist Standard or Enterprise license, or if your organization uses Gemini Code Assist for GitHub through Google Cloud, your access remain
45.
by kirtivr 5mo ago
+1 for this feature. And being able to download specific OSes instead of the entire 121GB zip file.
46.
by kirtivr 5mo ago
Love it, though it looks like the website got the HN hug of death. One of my favorite documentaries to learn the history of metal is "metal: a headbanger's journey" (available on YouTube).
47.
by kirtivr 5mo ago
Exactly. At the end of the day, we would rather have a more stable and bug-free kernel than not. It's not that much work for me anymore to report and even fix that obscure monitor driver bug that sometimes causes my machine to bootloop
48.
by kirtivr 5mo ago
I'd really like maintainers to get their hands dirty with AI agents as well to help speed up the reviews. Over the last year there have been way too many stories and Twitter posts like these. Yes, maintainers are overloaded, but that&#
49.
by kirtivr 5mo ago
A less known hack that can help people build professional websites that do not have the crazy gradients, wonky logos, typography and other tell-tale signs of being LLM generated: Get the TailwindCSS lifetime membership. Costs about $50 if I
50.
by kirtivr 5mo ago
Jetbrains does not have their own IDE-integrated coding agent? What do Jetbrains users use then? Amp?