It doesn’t. They run using stuff like Ollama or other LLM tools, all of the hobbyist ones are open source. All the model is is the inputs, node weights and connections, and outputs.
LLMs, or neural nets at large, are kind of a “black box” but there’s no actual code that gets executed from the model when you run them, it’s just processed by the host software based on the rules for how these work. The “black box” part is mostly because these are so complex we don’t actually know exactly what it is doing or how it output answers, only that it works to a degree. It’s a digital representation of analog brains.
People have also been doing a ton of hacking at it, retraining, and other modifications that would show anything like that if it could exist.
Yeah I’ve not had much of a deep dive into anything “ai” - closest thing I’ve got is a Google Coral monitoring my cameras. My current GPU selection is rather limited - sold my 1080ti and currently have a 3070 in my gaming rig + an un used 1660 which would run out of vram / be limited in what models they could run. Really not looking to run out and grab another card with more vram to play with either, maybe in a few years.
So the article is referring to the mobile app, and therefore would not have anything to do with someone running this model at home on their own hardware. I’ve not looked at DeepSeek’s repo yet but assuming there isn’t anything other than the model in the repo people need to calm down.
I run it locally on a 4080, how do they capture my keystrokes?
Are you running a quantized model or one of the distilled ones?
No idea, have you monitored the package / container for network activity?
Perhaps this refers to other clients not running the model locally.
It doesn’t. They run using stuff like Ollama or other LLM tools, all of the hobbyist ones are open source. All the model is is the inputs, node weights and connections, and outputs.
LLMs, or neural nets at large, are kind of a “black box” but there’s no actual code that gets executed from the model when you run them, it’s just processed by the host software based on the rules for how these work. The “black box” part is mostly because these are so complex we don’t actually know exactly what it is doing or how it output answers, only that it works to a degree. It’s a digital representation of analog brains.
People have also been doing a ton of hacking at it, retraining, and other modifications that would show anything like that if it could exist.
Yeah I’ve not had much of a deep dive into anything “ai” - closest thing I’ve got is a Google Coral monitoring my cameras. My current GPU selection is rather limited - sold my 1080ti and currently have a 3070 in my gaming rig + an un used 1660 which would run out of vram / be limited in what models they could run. Really not looking to run out and grab another card with more vram to play with either, maybe in a few years.
So the article is referring to the mobile app, and therefore would not have anything to do with someone running this model at home on their own hardware. I’ve not looked at DeepSeek’s repo yet but assuming there isn’t anything other than the model in the repo people need to calm down.
edit: deepstack to deepseek…
Yea, I am looking for a deep analysis of this as well
Are you looking for a deep learning of this as well