Scene_Cast2

7 ms·

91.
by Scene_Cast2 10mo ago
With what backhaul? WiFi?
92.
by Scene_Cast2 10mo ago
I'm not sure how likely it is that an answer would fall outside of the top-p of 0.95 (used in the paper). A random number generator would also need an unreasonably high number of samples to get a correct answer. I think figures 17 and
93.
by Scene_Cast2 10mo ago
I think my favorite of the bunch is the "Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model" paper. Easy to read, gets the point across very intuitively and quickly, and the point is ve
94.
by Scene_Cast2 10mo ago
Most likely the op amp (or whatever gain stage) noise. After a certain point, you get thermal noise. But with such scales, low sample rates and averaging are key.
95.
by Scene_Cast2 10mo ago
This post assumes C/C++ style business logic code. Anything HPC will benefit from thinking about how things map onto hardware (or, in case of SQL, onto data structures). I think way too few people use profilers. If your code is slow, p
96.
by Scene_Cast2 10mo ago
I found that taking a specific brand of Vitamin D (the Genestra D-mulsion in particular) right before bed was guaranteed to give me vivid dreams. I've had half a dozen friends try it, with every single one reporting similar results.
97.
by Scene_Cast2 10mo ago
Yep. MoE, FlashAttention, or sparse retrieval architectures for example.
98.
by Scene_Cast2 10mo ago
Still way pricier (>2x) than Gemini 3 and Grok 4. I've noticed that the latter two also perform better than Opus 4, so I've stopped using Opus.
99.
by Scene_Cast2 11mo ago
For anyone looking for some IDEs to tinker around with shaders: * shadertoy - in-browser, the most popular and easiest to get started with * Shadron - my personal preference due to ease of use and high capability, but a bit niche * SHADERed
100.
by Scene_Cast2 11mo ago
Matmuls (and GEMM) are a hardware-friendly way to stuff a lot of FLOPS into an operation. They also happen to be really useful as a constant-step discrete version of applying a mapping to a 1D scalar field. I've mentioned it before, bu
101.
by Scene_Cast2 11mo ago
My biggest worry is that it's harder and harder to find a phone with an unlockable bootloader.
102.
by Scene_Cast2 11mo ago
Super cool. Also, this is an example of why having an open OS is awesome.
103.
by Scene_Cast2 11mo ago
I've noticed that image models are particularly bad at modifying popular concepts in novel ways (way worse "generalization" than what I observe in language models).
104.
by Scene_Cast2 11mo ago
I would love to hear your (and others) opinions. I don't have a good idea of what happened inside or what they could have done differently, but I do remember them going from a world-leading LLM AI lab to selling embeddings to enterpris
105.
by Scene_Cast2 11mo ago
GPU horsepower isn't there for that.
106.
by Scene_Cast2 11mo ago
I swapped all my AA and AAA batteries for Enelooops. The cheaper white ones are the best for most applications.
107.
by Scene_Cast2 11mo ago
Where do you go for wide gamut prints? How do commercial printers compare to consumer printers in this regard? I'm working on a few wide gamut art pieces, and so far the test prints have been less than stellar. Disclaimer - I'm an
108.
by Scene_Cast2 11mo ago
I have 165Hz monitors. Software feels noticeably more snappy.
109.
by Scene_Cast2 11mo ago
I found out that in the embedded world (think microcontrollers without an MMU), Tensorflow lite is still the only game in town (pragmatically speaking) for vendor-supported hardware acceleration.
110.
by Scene_Cast2 11mo ago
OpenRouter has some gotchas with OpenAI models. In some cases it requires an OpenAI key.
111.
by Scene_Cast2 11mo ago
Theoretically yes, in practice no. There is (according to my sensors) a fairly large CO2 increase inside a room when a modern furnace (with external exhaust) is running. I've confirmed this with several units (all made in the last 10 y
112.
by Scene_Cast2 11mo ago
I was pushing back against "all kinds". The reason is that I've been seeing a number of inherently parallel architectures, but existing GPUs don't like some aspect of them (usually the memory access pattern).
113.
by Scene_Cast2 1y ago
There are large, large gaps of parallel stuff that GPUs can't do fast. Anything sparse (or even just shuffled) is one example. There are lots of architectures that are theoretically superior but aren't popular due to not being GPU
114.
by Scene_Cast2 1y ago
I noticed that they renamed the Element mobile app to Element Classic. Has Element X reached feature parity and stability yet? For how long will Classic be maintained?
115.
by Scene_Cast2 1y ago
If you need an attention sink, try chess! Pick a time control if it's over 2 minutes of waiting, and do puzzles if it's under. I find that there's not much of a context switch when I get back to work.
116.
by Scene_Cast2 1y ago
Running X11 on WSL is pretty easy and pain-free these days. The author could hypothetically run Zathura through that setup in a pinch. I'm all for a good Acrobat alternative though.
117.
by Scene_Cast2 1y ago
I wonder if there's research into fitting gaussian splats that are dependent on focus distance? Basically as a way of modeling bokeh - you'd feed the raw, unstacked shots and get a sharp-everywhere model back.
118.
by Scene_Cast2 1y ago
@dang sorry for the meta-comment, but why is yfontana's comment dead? I found it pretty insightful.
119.
by Scene_Cast2 1y ago
The juiciest one is the Albin Counter-gambit. If you follow the "ideal line" where white blunders and takes the bishop bait, there's a neat knight underpromotion to win a queen. From my own play, I typically see knight f3 fro
120.
by Scene_Cast2 1y ago
Are they competitive per Watt and per mm2 of die?

More ›