lambda

8 ms·

91.
by lambda 5mo ago
Some people would rather not hand over all of their ability to think to a single SaaS company that arbitrarily bans people, changes token limits, tweaks harnesses and prompts in ways that cause it to consume too many tokens, or too few to c
92.
by lambda 5mo ago
Their latest, Qwen3.6 35B-A3B is quite capable, and fast and small enough I don't really feel constrained running it locally. Some of the others that I've run that seem reasonably good, like Gemma 4 31B and Qwen3.5 122B-A10B just
93.
by lambda 5mo ago
It's a lighthearted, fun, visual benchmark that's not part of the standard benchmarks; and at least traditionally, it was not something that the labs trained on so it was something of a measure of how well the intelligence of the
94.
by lambda 5mo ago
Gah, the writing on this is so painful to read, it feels like this was most likely written by an LLM. The writing style is so unclear, it's hard to figure out one of the key points: it mentions that Gemini doesn't use a distinct u
95.
by lambda 5mo ago
The article has now been been semi-protected to prevent vandalism by anonymous users.
96.
by lambda 5mo ago
Or just the models are getting bigger and better at representing the long tail of the distribution. Previously errors like this would get averaged away more often; now they are capable of modelling more variation, and so are picking up on m
97.
by lambda 5mo ago
Or just because mistakes are part of the distribution that it's trained on? Usually the averaging effect of LLMs and top-k selection provides some pressure against this, but occasionally some mistake like this might rise up in probabil
98.
by lambda 5mo ago
I'm running an 8 bit quant right now, mostly for speed as memory bandwidth is the limiting factor and 8 bit quants generally lose very little compared to the full res, but also to save RAM. I'm still working on tweaking the settin
99.
by lambda 5mo ago
There is a certain amount of it which is the randomness of an LLM. You really want to ask most questions like this several times. That said, I have several local models I run on my laptop that I've asked this question to 10-20 times wh
100.
by lambda 5mo ago
I've run several local models that get this right. Qwen 3.5 122B-A10B gets this right, as does Gemma 4 31B. These are local models I'm running on my laptop GPU (Strix Halo, 128 GiB of unified RAM). And I've been using this co
101.
by lambda 5mo ago
Excuse me. Zero nines. Or two nines if you relax your definition of where they are in the number. https://infosec.exchange/@0xabad1dea/116334321751266751
102.
by lambda 5mo ago
You may be right that they were political in that sense. But also, they probably should have happened were he not a politician. He's been committing fraud and other white collar crimes for quite a while. Unfortunately, we go far too ea
103.
by lambda 5mo ago
Unfortunately, very few LLM benchmarks do this. LLMs get such high scores on many benchmarks because there's no difference between answering "I don't know" as giving a made up answer, and made up answers can improve the
104.
by lambda 5mo ago
I guess one of the things I don't understand: how you expect a stochastic model, sold as a proprietary SaaS, with a proprietary (though briefly leaked) client, is supposed to be predictable in its behavior. It seems like people are exp
105.
by lambda 6mo ago
While you can run with weights in RAM or even disk, it gets a lot slower; even though on any given token a fraction of the weights are used, that can change with each token, so there is a lot of traffic to transfer weights to the GPU, which
106.
by lambda 6mo ago
Yes, you can use it for local coding. Most harnesses can be pointed at a local endpoint which provides an OpenAI compatible API, though I've had some trouble using recent versions of Codex with llama.cpp due to an API incompatibility (
107.
by lambda 6mo ago
Right now, open models that run on hardware that costs under $5000 can get up to around the performance of Sonnet 3.7. Maybe a bit better on certain tasks if you fine tune them for that specific task or distill some reasoning ability from O
108.
by lambda 6mo ago
Well, one of the other reasons I suggest running it in a strictly limited container is that you can then run it in yolo mode. In fact, I use the pi agent, which doesn't have command sandboxing, it's always in yolo mode, I just r
109.
by lambda 6mo ago
Why do you expect that a weighted random text generator will ever behave in predictable way? How can people be so naive as to run something like Claude anywhere other than in a strictly locked down sandbox that has no access to anything but
110.
by lambda 6mo ago
Who would have guessed that running a binary blob dev tool, that is tied to a SaaS product, which was mostly vibe-coded, could lead to mysterious, hard to debug problems?
111.
by lambda 6mo ago
They aren't training new models for this. This is an agent harness for Opus 4.6.
112.
by lambda 6mo ago
It mostly trades some potential performance loss for speed, especially at longer contexts. Nemotron 3 Super doesn't perform quite as well on benchmarks as the similarly sized Qwen3.5 122B A10B model, but it goes faster and is cheaper t
113.
by lambda 6mo ago
Because something linear like Mamba doesn't perform as well; so you'd have a performance cliff, where suddenly the model would get more dumb and forget a lot of what was going on. Instead, you can get benefits from both by doing b
114.
by lambda 6mo ago
This is such a weird comment. Why do you think that "ethical vegans" like the "taste of plants" any more than anyone else? The whole point of being an ethical vegan/vegetarian is to not consume animals, not because
115.
by lambda 6mo ago
Where are you seeing dense? Most of the larger competitive models are sparse. Sure, the smaller models are dense, but over 30B it's pretty much all sparse MoE. And there are still plenty of hybrid architectures. Nemotron 3 Super 120B
116.
by lambda 6mo ago
Just because Facebook supports it doesn't mean it's bad. They may not support it for the same reasons, they probably just don't want the cost and liability of doing identify verification themselves and so want to make sure al
117.
by lambda 6mo ago
Slippery slope fallacy. We shouldn't object to a reasonable law just becasue it might, theoretically, pave the way to an unreasonable law. In fact, this is put in place as an alternative to the kind of law being enacted elsewhere, righ
118.
by lambda 6mo ago
Here's the law: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtm... It requires that operating systems provide a way, at account setup, to specify the age or birthdate of a user, and provides an API fo
119.
by lambda 6mo ago
They are not legislating specific APIs. They are legislating that an API has to be provided, just like other laws legislate that you have to provide accessibility APIs, but the details of the APIs are left up to the companies. I work in avi
120.
by lambda 6mo ago
The whole point of the California/Colorado laws is to provide an alternative to that. The whole point is that it provides a privacy preserving way to provide a signal about whether someone is in a particular age bracket, without requir

More ›