Viewing profile — juliangoldsmith

The claim isn't so wild when it's a generalist versus a finetune trained specifically on the tasks being benchmarked.

That benchmark ranks Kimi K2.6 and K2.7 Code near the bottom. Both are below Ornith 35B. It ranks Gemma 4 26B much higher than GLM-5.2. The results don't make much sense.

It looks like they're comparing Orinth 9B to Qwen 3.5 35B, not Qwen 3.6. I guess it kind of makes sense since it's a finetune of 3.5, but I totally missed until I looked closely. I…

It should improve agents' action selection by allowing them to evaluate actions' effects before performing them. An agent using only a regular LLM has no real way to predict the re…

One can only imagine how much money Intel would have made from Optane during the ongoing RAM shortages. It would be absolutely perfect for warm KV cache, and potentially good for M…

No true Scotsman would ever use binary blobs.

Blocking by age rating takes out the majority of the classic Disney movies and shows. They only consider the newer CGI stuff "child-friendly".

The silenced errors aren't guaranteed to be memory leaks or use after frees. There are some situations where memory is being handled properly, but the borrow checker isn't able to …

It sounds like you were stuck between a rock and a hard place there. Hope the Rust integration goes well.

>trading off for speed If speed is a concern, why did you all stick with Synapse (essentially single-threaded due to the GIL) over moving to Dendrite? As far as I can tell, Dendrit…

It doesn't appear to be open source, so users have no control or lasting guarantees of privacy.

As someone who as attempted to use React Native for Windows, I can tell you that the "native" XAML doesn't make things any better. If it was using web technologies I wouldn't need …

"Hard Refresh" and "Airglow" made it onto my "On Repeat" playlist almost immediately.

I'd use caution with the Mi50s. I bought a 16GB one on eBay a while back and it's been completely unusable. It seems to be a Radeon VII on an Mi50 board, which should technically w…

I've been using Azure AI Foundry for an ongoing project, and have been extremely dissatisfied. The first issue I ran into was with them not supporting LLaMA for tool calls. Microso…

What is it that makes higher order derivatives less useful at high dimensionality? Is it related to the Curse of Dimensionality, or maybe something like exploding gradients at high…

It works out of the box without jumping through any hoops, and the fact that it has an OpenCL backend means it can run on a wide variety of hardware. I don't know of any other auto…

If we were dealing with pure cosmic background radiation, or inside a Faraday cage, sure. What I'm referring to are things like radio broadcasts, 60 Hz hum from power lines, noise …

How does Tinygrad fall short? Performance is fine [0]. It's much smaller than Pytorch and all, but that's kind of in the name. I've been hearing about MLIR and OpenXLA for years th…

AMD's hardware might be compelling if it had good software support, but it doesn't. CUDA regularly breaks when I try to use Tensorflow on NVIDIA hardware already. Running a poorly-…

>How does the disconnected audio input of any random PC or thinclient compare? That will give you RF noise, which isn't really random.

Most debit cards in the US can be run either as debit or credit. Debit transactions require a PIN, but credit transactions don't.

The search engine they linked to happens to provide a significant portion of Mozilla's revenue.

That Wikipedia page explicitly states that the public has never subsidized damage from a nuclear accident. Nuclear energy companies are required to have $450M in private insurance …

Rank in the array language terminology seems to be synonymous with tensor order (sometimes called tensor rank). Ken Iverson (APL's creator) was a mathematician, so I'm not that sur…