icelancer

6 ms·

1.
by icelancer 4d ago
Huh. I never considered this. Hacking TVs with firmware would be a great project for LLMs to fuzz.
2.
by icelancer 7d ago
I've played both. The GM plays tremendously differently than Stockfish. Engines - specifically heuristically-driven ones like Stockfish - don't play like a strong GM. They play engine-perfect chess, which isn't how a GM plays
3.
Vast.ai Benchmarking – Detect/Log Garbage Servers(github.com)
2 points by icelancer 24d ago|0 comments
4.
by icelancer 1mo ago
I think the point is that a larger model with far less reasoning time solves the problem just fine. Which of course is the tradeoff: The smaller the model, the more reasoning you need to get decent answers to tough questions.
5.
by icelancer 1mo ago
> I can imagine a curve where for a given level of “intelligence” you either need model size, or inference time (“test time compute”), and can somewhat trade one for the other. This is exactly what was shown on Luna/Terra/Sol t
6.
by icelancer 1mo ago
it does this at a variable rate of frames which you can set - not sure if it is transcribing or natively understanding audio, but I think it's the latter since it is much faster than most transcription models I am aware of Regardless y
7.
by icelancer 1mo ago
Crazy it's still the only video understanding endpoint. It's what I use it for and no other model even offers a competitor.
8.
by icelancer 1mo ago
Great point. The EA group have a ton of parallels to the Scientologists, so all of this is not surprising.
9.
by icelancer 1mo ago
I use Modal for fine tuning and unsloth mostly.
10.
by icelancer 2mo ago
This is what I've been waiting for. We are still using fine-tuned deployments of Qwen3.6-27B with a lot of success but could use a bump in intelligence. Here's hoping.
11.
by icelancer 2mo ago
This is actually a very good question. The funniest and simplest answer is that you make money off yourself.
12.
by icelancer 2mo ago
I asked an adjacent question to a broker friend who flatly replied: "Not interested in being served a Wells Notice." Which more or less answered the question for me, even if he was being a bit hyperbolic.
13.
by icelancer 2mo ago
Gets a 100% on Pangram. Stuff is so distracting. Write your own posts, FFS. Or at least pass it through "humanizer" type plugins.
14.
by icelancer 2mo ago
dash dash "--" on a lot of systems and word processors turns it into the em-dash automatically
15.
by icelancer 4mo ago
repo is cloned, AGENTS.md is auto-read into context, the doc says to not allow PR spam. think of it like a soft prompt hook.
16.
by icelancer 5mo ago
This is only true for some parts of the time cost function.
17.
by icelancer 5mo ago
love this adversarial work
18.
MoA-X: Mixture of Agents Orchestration Framework(github.com)
2 points by icelancer 5mo ago|0 comments
19.
by icelancer 5mo ago
Lambda calculus?!
20.
by icelancer 5mo ago
Handing over email metadata, or whatever your interrogator wants from you, will only cause them to shift the goalposts, or find something they want to find in the metadata even if it exonerates you. There is no reason to cooperate with jour
21.
by icelancer 5mo ago
All of us "olds" had this as a hobby. Or used it regularly. Just a bunch of weird stuff in this article.
22.
by icelancer 6mo ago
No doubt. Especially for GPU inference at scale. We overclock/overvolt for training and tune way down for inference.
23.
by icelancer 6mo ago
Self-racking lets you rack a bunch of gear you'd never find in VM/dedicated rentals, like consumer parts or older, still very good parts. Overclocking options are available as well if you DIY. If you need single-threaded performan
24.
by icelancer 7mo ago
> What am I missing? VRAM capacity given the Cerebras/Groq architecture compared to Nvidia. In parallel, RAM contracts that Nvidia has negotiated well into the future that other manufacturers have been unable to secure.
25.
by icelancer 8mo ago
Interesting - I'll try some other methods
26.
by icelancer 8mo ago
I generally agree with you. But I put up with it since Notepad++ is good software. It is what it is.
27.
by icelancer 8mo ago
I've had a lot of problems with even 10GbE via Thunderbolt 3/4. Bandwidth-wise it works fine, but latency and jitter are issues. This means that stuff like high-speed cameras that need to be synchronized over Ethernet using Precis
28.
by icelancer 8mo ago
Yeah, Notepad++ is known for political messaging in their updates. Taiwan, Ukraine, etc.
29.
by icelancer 8mo ago
Their compaction endpoint is far and away the best in the industry. Claude's has to be dead last.
30.
by icelancer 8mo ago
> sub 900 not possible though. I performed a similar analysis to you and found it very difficult to imagine sub-1000. Your comment I think convinced me that it may be possible, though. Interesting. I'm below the threshold for recrui

More ›