broretore

Joined 11 months ago

Karma

4

Submissions

18

Member since

Dec 9

View on Hacker News

Recent submissions

Comment · 10 months ago

would the world be worse off if instead of google it had been blooglie or hooli that succeeded?

i don't know

Comment · 10 months ago

Do you want to fix the typo?

Comment · 10 months ago

any time i am certain of something i never capitalize and i do not end my sentences with periods or use any punctuation because i like people to believe i am an omniscient narrator who cannot be interrupted

Comment · 10 months ago

Comment · 10 months ago

Both Udemy and Coursera are public companies?

Comment · 10 months ago

Wow, what a blast from the past -- I went to high school with a guy named Erez Hadari.

Comment · 10 months ago

You could also epoxy it.

Comment · 10 months ago

This is basically a tinybox pro?

Comment · 11 months ago

Telling ChatGPT to do creative writing for you isn't creative writing ser.

Comment · 11 months ago

Comment · 11 months ago

> It started, as many things do these days, by scrolling on X.

Thanks Blake. This is the sign I needed to cancel my Twitter subscription and hellban X from my LAN.

Comment · 11 months ago

Facts don't matter when Elon's feelings are at stake

Comment · 11 months ago

How do you write nine paragraphs without once checking the repo for code, or noticing the obvious Grok confabulations throughout the paper?

This should concern you. The next person to get LLM psychosis might be you.

Comment · 11 months ago

Your paper is 10 pages of fluff without even an architecture diagram or a single equation, bro. It's not real.

Comment · 11 months ago

CIFAR-10 is an image classification dataset (32x32 pixel images.

LLaMA 70B 3.3 is a text-only, non-multimodal language model. Just look up the Huggingface page that your own repo points to.

> The Llama 3.3 instruction tuned text only model...

I might be wrong, but I'm pretty sure a text model is going to be no better than chance at classifying images.

Another comment pointed out that your test suite cheats slightly on HellaSwag. It doesn't seem unlikely that Grok set up the project so it could cheat at the other benchmarks, too.

https://news.ycombinator.com/item?id=46215166

> The repo contains the full pipelines, configuration files, and benchmark scripts, and those show the precise datasets, metrics, and evaluation flows.

There's nothing there, really.

I'm sorry that Grok/Ani lied to you, I blame Elon, but this just doesn't hold up.