I think some people do like the collection mechanics. And some games require specific team comps in order to beat the harder content.

How do you determine that the model was reasoning in Chinese in layer X? I would think the middle layers do not map into any tokens.

Using a logit lens (prior art: https://www.lesswrong.com/posts/AcKRB8wDpdaN6v6ru/interpreti...)

Not necessarily. Servers serving the model likely has enough traffic that they are batching decodes already. MTP reduces latency and increase efficiency only when the server can’t batch enough concurrent streams to be compute bound rather than memory bound.

Fair didn't think about batching makes more sense for self hosted models then.

Running the mxfp4 unsloth quant of qwen3.5-397b-a17b, I get 40 tps prefill, 20tps decode.

AMD threadripper pro 9965WX, 256gb ddr5 5600, rtx 4090.

Funny, when I got tired of trying to find a nice desktop background I just started using a solid color of muted blue or green. I never read about this specific usage of colors before but I bet I saw something somewhere that clued me in on this color.

Reminding me that early Windows versions used to have this colour as the default desktop colour -- and I remember seeing similar tones on Mac and Unix desktops in the 90s.

https://faculty.washington.edu/ajko/books/user-interface-sof...

I tried fedora silverblue for a while, but the way it works is that it builds a new root fs image whenever you change the installed packages, this makes system package changes take comparatively long vs a traditional os. They suggest installing most apps via flatpak, which is okay as long as you can deal with flatpak idiosyncrasies.

I also tried fedora coreos for a vm + container host, but found the recommended method to configure the system with ignition files and one shot systemd units to be too involved for making a one off system, and it’s probably better for a cloud deployment with many identical nodes.

In all fairness, Nix is similarly slow.

I think it’s not so much that the asyncio primitives got wrong about shared state, as much as is what the authors got wrong about the usage of those primitives. They are classic concurrency primitives that’s been around for almost half a century. They work as designed, but require some care to use correctly.

Agreed. This isn't an asyncio problem, it's just not how those primatices work.

I noticed requests that were exploiting the vulnerability were turning into timeouts pretty much immediately after rolling out the patch. I’m surprised it took so long for it to be announced.

Is there a mypy plugin or other tool to check this via static analysis before runtime? To my knowledge jaxtyping can only be checked at runtime.

I doubt it, since jaxtyping supports some quite advanced stuff:

    def full(size: int, fill: float) -> Float[Array, "{size}"]:
        return jax.numpy.full((size,), fill)

    class SomeClass:
        some_value = 5

        def full(self, fill: float) -> Float[Array, "{self.some_value}+3"]:
            return jax.numpy.full((self.some_value + 3,), fill)