juliangoldsmith

6 ms·

1.
by juliangoldsmith 1mo ago
Impulse control issues, like ADHD, can factor in too.
2.
by juliangoldsmith 1mo ago
A possible "third variable" could be intelligence. It's been shown to be strongly correlated with success in general, and it seems likely people of similar intelligence would find it easier to relate to each other.
3.
by juliangoldsmith 3mo ago
The claim isn't so wild when it's a generalist versus a finetune trained specifically on the tasks being benchmarked.
4.
by juliangoldsmith 3mo ago
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.
5.
by juliangoldsmith 3mo ago
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. In my brief tests, Ornith 35B performed quite well
6.
by juliangoldsmith 3mo ago
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 results of its actions. It has to just take an acti
7.
by juliangoldsmith 5mo ago
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 MoE expert offloading.
8.
by juliangoldsmith 5mo ago
No true Scotsman would ever use binary blobs.
9.
by juliangoldsmith 9mo ago
Blocking by age rating takes out the majority of the classic Disney movies and shows. They only consider the newer CGI stuff "child-friendly".
10.
by juliangoldsmith 9mo ago
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 prove it. One example might be a tree-like struct
11.
by juliangoldsmith 9mo ago
It sounds like you were stuck between a rock and a hard place there. Hope the Rust integration goes well.
12.
by juliangoldsmith 9mo ago
>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, Dendrite is, for all intents and purposes, abandoned.
13.
by juliangoldsmith 9mo ago
It doesn't appear to be open source, so users have no control or lasting guarantees of privacy.
14.
by juliangoldsmith 10mo ago
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 to manually modify RNSVG to fix segfaul
15.
by juliangoldsmith 11mo ago
"Hard Refresh" and "Airglow" made it onto my "On Repeat" playlist almost immediately.
16.
by juliangoldsmith 1y ago
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 work. It immediately hangs the first time an Open
17.
by juliangoldsmith 1y ago
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. Microsoft stated in February that they were working on it [0]
18.
by juliangoldsmith 2y ago
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 higher orders?
19.
by juliangoldsmith 2y ago
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 autograd libraries with a non-CUDA backend, but I'd be
20.
by juliangoldsmith 2y ago
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 put out by switching power supplies, and that sort of t
21.
by juliangoldsmith 2y ago
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 through Tensorflow, but I've never seen
22.
by juliangoldsmith 2y ago
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-implemented clone of CUDA where even getting Pyt
23.
by juliangoldsmith 2y ago
>How does the disconnected audio input of any random PC or thinclient compare? That will give you RF noise, which isn't really random.
24.
by juliangoldsmith 2y ago
Most debit cards in the US can be run either as debit or credit. Debit transactions require a PIN, but credit transactions don't.
25.
by juliangoldsmith 2y ago
The search engine they linked to happens to provide a significant portion of Mozilla's revenue.
26.
by juliangoldsmith 2y ago
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 for each reactor. For amounts over that, the Price-Anderso
27.
by juliangoldsmith 2y ago
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 surprised to learn the term came from a branch of ma
28.
by juliangoldsmith 2y ago
The array languages aren't super popular because of the sharp learning curve. They're a lot different than most other languages, and they have a lot of operators that simply don't exist in something like C++. A few years ago
29.
by juliangoldsmith 2y ago
The title is somewhat misleading, since it implies the family paid that much. From near the bottom of the article: >Brigland’s family paid $7,200, their plan’s out-of-pocket maximum.
30.
by juliangoldsmith 2y ago
Dupe: https://news.ycombinator.com/item?id=42010136

More ›