5555watch

7 ms·

1.
by 5555watch 4d ago
Recalling their recent misuse report [0], I kind of agree with them. And performance/intelligence wise I think we are already at quite a nice local optima. [0]: https://www.anthropic.com/threat-intelligence-report-septe
2.
by 5555watch 6d ago
This reminded me of anecdotes of people discussing with friends about buying a random specific item, and then suddenly seeing it advertised everywhere before even googling about it. Next step, discussing your Navier Stokes solutions with fr
3.
OpenAI puts Pro subscriptions on hold due to Astra demand(techcrunch.com)
7 points by 5555watch 6d ago|0 comments
4.
by 5555watch 6d ago
At least in my experience, the issue with ArXiv is that the expectation is that the draft should be already in a good enough state. And polishing plus writing the meat around the main result can take a lot of time
5.
by 5555watch 6d ago
Does it make sense to start privately and then open the repo after publication? Will the dates be retained? Also, isn't commit history easy to spoof?
6.
by 5555watch 6d ago
Is there an alternative link without certificate issues?
7.
by 5555watch 6d ago
I also think that it's quite a bad PR for them, is it really worth the Millenium prize? Is it not enough that top mathematicians are already actively using these tools? In the long term this would lead to potentially profitable collabo
8.
by 5555watch 6d ago
The ultimate drive for some researches is the pursuit of knowledge. If I'm stuck at some block which prevents me from continuing in some direction that I want, of course I would like some help. I believe we already have nonzero collabo
9.
by 5555watch 7d ago
Or, if you're using, you need to push a pre-print relatively soon after AI help.
10.
by 5555watch 8d ago
Maybe a dumb observation, but if a chain of people were working on the problem for a long time, it's not difficult to imagine that someone accidentally prompted a model with their personal or some other account without the privacy set
11.
by 5555watch 13d ago
Citing Tibo [0]: " - Bigger number = Better - Bigger celestial object = Better and the scale is Astra > Sol > Terra > Luna. " [0]: https://x.com/thsottiaux/status/2095600295808283073
12.
by 5555watch 13d ago
Many other links are also 404'ing, so probably internal issues. https://openai.com/index/legora-financial-statement-review-w... https://openai.com/index/playco-game-prototyping-with-astra&#x2F
13.
by 5555watch 13d ago
Probably this complexity was needed to beat all those benchmarks.. While I hate the code it produces, and the overwhelming documentation, I really enjoy how sometimes it's able to keep trying new things and testing, till it finds somet
14.
by 5555watch 15d ago
It means the questions including their answers are dependent. Ie, theres a data generating process for them, that the model uncovers. Like a KNN is known to have near Bayes accuracy as k/n to 0, n to infty, k to infty. The data reveals
15.
by 5555watch 15d ago
It makes sense. Even if it finds some exploit on your own code, who's to say you can't reuse the same exploit on some other system?
16.
by 5555watch 15d ago
While I can't speak for everyone in academia, I personally don't feel comfortable in putting my research questions and outputs to a private website, before the idea is at least arxived. Especially as all the Fable/Mythos prom
17.
by 5555watch 15d ago
Is Fable 5.1 still actively downthrottling the reasoning when questions relate to frontier ML questions, like it did with 5.0?
18.
by 5555watch 16d ago
It's not clear to me why isn't the 5.6 Sol Pro available on Codex. Maybe due to the high computational demands and slow speed. But as an anecdote, once when the Sol Ultra models were stumped on one problem, I gave the full extende
19.
by 5555watch 26d ago
I can't be the only one that scrolls past quickly the whatever Google-AI spurts out. If I wanted to ask an AI - I would. But I've learned that they hallucinate stuff with numbers or anything that's time sensitive/recent
20.
by 5555watch 28d ago
>Everyone should occasionally go back to the old models to see how much worse they were, like even a year ago you could generate results but they were typically full of bugs Oh yeah, I'm still amazed how good the current iteration o
21.
by 5555watch 1mo ago
All of your use cases are very advanced. I recently used it at grocery stores in a foreign country. Photographed the whole aisle and told it to find Y (detergent, softener, glue, sour cream, whatever), at the same time recommend the best Y
22.
by 5555watch 1mo ago
But the words don't really matter, do they? The model thought "user said believe in yourself, it means they want me to continue".. AI did a fixed amount of guesses, didn't yield anything. It probably documented the tries
23.
by 5555watch 1mo ago
PCA is good, but you could also try playing around with Sparse (robust) PCA. The sparsification loses orthogonality, but does not necessarily lose information, it can yield a different rotation and cleaner vectors. Now whether that matters
24.
by 5555watch 1mo ago
Even for consumers.. Every enthusiast now wants those specced out rigs to play with LLMs. Make a nice chip for that, and it will reduce some pressure on consumer RAM demand
25.
by 5555watch 1mo ago
Is it still true? I'd assume you should be able to freeze the matrix and unfreeze an expert block, before feeding particularly chosen training data. Or that doesn't work?
26.
by 5555watch 1mo ago
In my understanding the first Deep Think / Pro models were already very good as they were doing some kind of parallel repeated reasoning, thus were slow and expensive. So if chatjimmy speeds enables a fast deep think level performance,
27.
by 5555watch 1mo ago
I mean in their code the simulation is incorrect. Someone found the source: https://github.com/pem725/Dunning-Kruger They're sampling slope and bias for the line from U[0,1] and U[0,100], the expected values of wh
28.
by 5555watch 1mo ago
It's an interesting read. Curiously, it doesn't really debunk anything. The fact that given X and Y random and independent, that Y-X is correlated with X doesn't disprove the Dunning Kruger. It in fact proves that Y = 1 X is
29.
by 5555watch 1mo ago
The extrapolation can also be a learned skill, especially in math. How many papers took result X, extended it to Y using known building blocks, and applied to Z. By the way, convex hull permits extrapolating past the training data. LLM won&
30.
by 5555watch 1mo ago
They're simulating randomness incorrectly: relationship between true and perceived will average 0.5, not 0; and bias will average 50%, not 0%. That's why their "random data" is sloped. Add negative relationship and negat

More ›