GaggiX

27 ms·

571.
by GaggiX 2y ago
The larger GPT-2s were also considered too dangerous to release publicly at first.
572.
by GaggiX 2y ago
>our latest large-scale dataset, encompassing up to 18 trillion tokens I remember when GPT-3 was trained on 300B tokens.
573.
by GaggiX 2y ago
I don't think a single 100-word email using GPT-4 alone can cost 140wh, even considering how expensive GPT-4 is to run. If it takes 20 seconds for the model to compose the letter, that means: (3600/20)*140=25200W, a 25200W piece o
574.
by GaggiX 2y ago
On my phone it has some extreme borders and in many lines it can only fit one single word.
575.
by GaggiX 2y ago
Just seeing how bad the site renders on my phone I wouldn't trust it.
576.
by GaggiX 2y ago
This seems the usual CoT that has been used for a while, o1 was trained with reinforcement learning with some unknown policy, so it's much better at utilizing the chain of thought.
577.
by GaggiX 2y ago
I wonder why it wasn't mentioned in the article.
578.
by GaggiX 2y ago
Mr.Beast is in some big controversies right now, and it's honestly much more interesting than this PDF, I expected to see the "no does not mean no" section in this PDF.
579.
by GaggiX 2y ago
I'm very optimistic about it because native multimodal LLMs have hardly been explored. Also in general, I have yet to see these models plateau, Claude 3.5 Sonnet is a day and night different compared to previous models.
580.
by GaggiX 2y ago
Just remember to never remove the stolen iPhone from Find My, the thieves will try everything to convince you to do so.
581.
by GaggiX 2y ago
The takeaway is also that o1-preview is a major improvement compare to GPT-4o. Anthropic is just ahead.
582.
by GaggiX 2y ago
It really shows how far ahead Anthropic is/was when they released Claude 3.5 Sonnet. That being said, the ARC-agi test is mostly a visual test that would be much easier to beat when these models will truly be multimodal (not just appen
583.
by GaggiX 2y ago
Do you have a source about OpenAI hiring a few hundred professors, journalists and writers? Because I honestly doubt.
584.
by GaggiX 2y ago
>the 'big leap forward' is just another dataset of people repairing chatgpt's lack of reasoning capabilities. I think there is a really strong reinforcement learning component with the training of this model and how it has
585.
by GaggiX 2y ago
Well they did in the testing I have seen.
586.
by GaggiX 2y ago
Yeah, a solution to the counting letters problem would probably be encoding the letters into the large embeddings by probably adding a really small loss and a small classification model on the embeddings so that it would learn to encode the
587.
by GaggiX 2y ago
I imagine that GPT-5 would be a refined version of this paradigm, probably with omni (multimodal) capabilities added (input and output).
588.
by GaggiX 2y ago
The website seems to redirect me to a low resolution image, the first time I clicked on the link it worked as you are saying.
589.
by GaggiX 2y ago
Did you edit the message? I cannot see anything now in the screenshot, too low resolution
590.
by GaggiX 2y ago
Have you tried Claude 3.5 Sonnet?
591.
by GaggiX 2y ago
I didn't know about NotebookLM, it seems like a very useful tool.
592.
by GaggiX 2y ago
Yes, but people would probably not realize that it was made with significant assistance from AI.
593.
by GaggiX 2y ago
Exactly, https://arxiv.org/abs/2207.12598 Jonathan Ho, one of the authors of the CFG paper, now works for Ideogram, and Ideogram 2 is one of the very few models (or perhaps the only one) where I don't see the arti
594.
by GaggiX 2y ago
>The days of free web scraping especially for the richer sources of material are almost gone I would say the opposite, it has never been easier to collect a huge amount of data, in particular if you have a target, also you don't eve
595.
by GaggiX 2y ago
The style becoming identifiable and stale has mostly to do with CFG and almost nothing with the dataset, the heavy use of CFG by most models trades diversity with coherency. You don't need a costant intake of new images and styles, it&
596.
by GaggiX 2y ago
There are much newer papers shown than "Attention is All You Need" (all of them?) and much less talked about (probably all of them, too). It shouldn't be surprising that a LLM is able to understand a paper, just upload one to
597.
by GaggiX 2y ago
Did they removed the book section? I can only find the "papers" section now.
598.
by GaggiX 2y ago
A very useful guide about how diffusion models work and implementation: https://keras.io/examples/generative/ddim/ I find the explanation in this article very intuitive.
599.
by GaggiX 2y ago
I read about the Rohonc Codex and it seems that most people agree that it's a hoax, the language itself doesn't seem to share the statistical property of a real language it seems, Rohonc Codex is the level of sophistication that I
600.
by GaggiX 2y ago
Everything about this manuscript is so mysterious, the origins, the language, the plants and other drawings; still not mysterious enough that you would that it would be impossible to find a solution to all of this, I can understand the pass

More ›