Non-blogspam link: https://qwenlm.github.io/blog/qwq-32b-preview/

discussion: https://news.ycombinator.com/item?id=42259184

I wish I knew that there existed a public quantum AI company as an easy short before they went out of business.

unfortunately the company can appear to be in either state until it collapses

That's known as the Heisenberg Uncertainty Principle of tech companies.

Counter-example: IONQ Inc whose stock jumped up after a recent military contract.

See also Rigetti Computing

The model page says the prompt is:

The system prompt used for training this model is:

   You are a world-class AI system, capable of complex reasoning and reflection. Reason through the query inside <thinking> tags, and then provide your final response inside <output> tags. If you detect that you made a mistake in your reasoning at any point, correct yourself inside <reflection> tags.

from: https://huggingface.co/mattshumer/Reflection-70B

But their model is fine-tuned on top of Llama, other base models could not follow this specific system prompt.

Nice! It would be a better benchmark to compare this prompt (w/ gpt-4o, claude) with whatever the original model was compared to.

It appears that they have. [0]

[0] https://x.com/mattshumer_/status/1831767017507954808

Google Colab has this, I wouldn't be surprised if there was a Jupyter widget to implement something similar.

Edit: looks like Mercury (A jupyter extension) has them: https://runmercury.com/docs/input-widgets/

Title says "for everyone", but post says "Phind-405B is available now for all Phind Pro users". I guess everyone on earth has paid for Phind :)

The "for everyone" part is about the new Phind Instant, trained using similar data to Phind-405B, which is great at fast information retrieval

Are you a VC? If they really didn't care about their investment exits, that would be crazy.

It’s not that they don’t care, of course they want to find winners. It’s just that A) there is so much capital to allocate that they have to allocate to marginal ideas B) their priorities are to raise their next fund which means focusing on vanity metrics like IRR and paper markups C) The incentive structure in VC pushes them to invested based on motivated reasoning. Remember VC returns are cyclical, and many vintages underperform the public markets and particularly large funds do worse simply because they have too much capital to allocate and too few great ideas.

I think they treat success and failure as mostly luck, with a tiny bit of hygiene

The trick is being in the game long enough

And you’d be wrong of course, like any other random guess you could make about a topic you know nothing about.

Can run on CPU (slower) or AMD GPUs.

What about Mac/Metal?

this is what I was getting at, i tried on my mbp and no luck. might be just an installer issue but I wanted confirmation from someone with more know-how before diving in

I got SAM 1 to work with MPS device on my MacBook Pro M1, don’t know if it works with this one too.

Might have issues if you're from Texas or Illinois due to their local laws.

What is the Illinois law?

Edit: Found lower in thread: biometric privacy laws

Video games, for one.