That feels like a problem with your harness. What are you using it with?

You forgot to mention your tool does worse for 8/16 LLMs compared to replace?

Problem is, replace has been around for so long, most LLMs are tuned for it now

This is gorgeous. I see a install on pwa, but it would be even nicer if this could be an android widget. Anyone know if this is possible?

it's a pwa and the widgets are a good idea

Arc agi was never a good benchmark that tested spatial understanding more than reasoning. I'm glad it's no longer popular

What do you mean? It definitely tests reasoning as well, and if anything, I expect spatial and embodied reasoning to become more important in the coming years, as AI agents will be expected to take on more real world tasks.

spatial or not, arc-agi is the only test that correlates to my impression with my coding requests

Pro isn't even a model. If they actually used a model name I'd think they were just into LLMs. Chatgpt pro is a specific paid service.

Yes, ChatGPT Pro is a paid service, but FWIW I give OpenAI money. They're not giving me money.

Great explanation for why AI slop is so distracting. It's actively hurting my ability to find high quality things, because it's a strong signal for low effort, and yet so many good projects are starting to use it on readmes.

I have no idea if OPs project is good, but the idea is.

Compared to?

Vivaldi still has it as an option. That said, I keep them combined and can't remember the last time it guessed wrong.

Odd you should say that, as that's exactly what they've been discussing

No it's not. This article is about proposing Brotli as another possible '/Filter' for stream objects, like content streams (page drawing commands). Images are streams too, but unless you mean compressing raw pixel bytes in Brotli, there's no mention of a JPEG-XL or WEBP filter.

well, not mentioned in this specific article. But JPEG-XL support is something they're working on [1].

[1]: https://pdfa.org/wp-content/uploads/2025/10/PDFDays2025-Brea...

Oh cool!! TIL

Not all. In fact a small model that has none of them but loads them on demand might be the most efficient thing