Hacker News: rain1

CleoBench: Can Fable mathematically prove Cleo's integrals?

rain1 — Fri, 10 Jul 2026 07:43:23 +0000

Article URL: https://rain-1.github.io/cleo-bench/

Comments URL: https://news.ycombinator.com/item?id=48856938

Points: 2

# Comments: 0

New comment by rain1 in "Transformer Architecture Visualizer"

rain1 — Sat, 13 Dec 2025 11:04:48 +0000

I've used Google Antigravity to write scripts to download and produce architecture diagrams for various LLMs from huggingface. It's pretty useful so I thought I'd share it.

There's also a model comparison spreadsheet that you can compare sizes and such https://weavers.neocities.org/architecture-encyclopedia/mode...

If you'd like any additional models to be added I can add them in.

Transformer Architecture Visualizer

rain1 — Sat, 13 Dec 2025 11:04:48 +0000

Article URL: https://weavers.neocities.org/architecture-encyclopedia/

Comments URL: https://news.ycombinator.com/item?id=46253723

Points: 2

# Comments: 1

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 15:08:49 +0000

The Gemma models are too small to be included in this list.

You're right the T5 stuff is very important historically but they're below 11B and I don't have much to say about them. Definitely a very interesting and important set of models though.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 15:05:41 +0000

Yes but just purely in terms of entropy, you can't make a model better than GPT-4 by training it on GPT-4 outputs. The limit you would converge towards is GPT-4.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 13:08:13 +0000

This is kind of related to the jack morris post https://blog.jxmo.io/p/there-are-no-new-ideas-in-ai-only he discusses how the big leaps in LLMs have mostly come - not so much from new training methods or arch. changes as such - but the ability of new archs. to ingest more data.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 13:06:46 +0000

It's extremely interesting how powerful a language model is at compression.

When you train it to be an assistant model, it's better at compressing assistant transcripts than it is general text.

There is an eval which I have a lot of interested in and respect for https://huggingface.co/spaces/Jellyfish042/UncheatableEval called UncheatableEval, which tests how good of a language model an LLM is by applying it on a range of compression tasks.

This task is essentially impossible to 'cheat'. Compression is a benchmark you cannot game!

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 13:05:17 +0000

I think that one thing that this chart makes visually very clear is the point I about GPT-3 being such a huge leap, and there being a long gap before anybody was able to match it.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 13:03:33 +0000

This is really awesome. Thank you for creating that. I included a screenshot and link to the chart with credit to you in a comment to my post.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 12:16:53 +0000

I can correct mistakes.

> it somehow merged Llama 4 Maverick's custom Arena chatbot version with Behemoth

I can clarify this part. I wrote 'There was a scandal as facebook decided to mislead people by gaming the lmarena benchmark site - they served one version of llama-4 there and released a different model' which is true.

But it is inside the section about the llama 4 model behemoth. So I see how that could be confusing/misleading.

I could restructure that section a little to improve it.

> Llama 405B was also trained on more than 15 trillion tokens[1],

You're talking about Llama 405B instruct, I'm talking about Llama 405B base. Of course the instruct model has been traiend on more tokens.

> why is there such a focus on token training count?

I tried to include the rough training token count for each model I wrote about - plus additional details about training data mixture if available. Training data is an important part of an LLM.

New comment by rain1 in "How large are large language models?"

rain1 — Wed, 02 Jul 2025 11:55:26 +0000

I have corrected that. It was supposed to say "None of this document was written by AI."

Thank you for spotting the error.

How large are large language models?

rain1 — Wed, 02 Jul 2025 10:39:38 +0000

Article URL: https://gist.github.com/rain-1/cf0419958250d15893d8873682492c3e

Comments URL: https://news.ycombinator.com/item?id=44442072

Points: 263

# Comments: 150

Midjourney Generating Screenshots of Movies

rain1 — Wed, 18 Sep 2024 15:35:04 +0000

Article URL: https://unrollnow.com/status/1835752715537826134

Comments URL: https://news.ycombinator.com/item?id=41581095

Points: 2

# Comments: 0

New comment by rain1 in "Tell HN: Burnout is bad to your brain, take care"

rain1 — Fri, 06 Sep 2024 07:48:23 +0000

> Take care of your mental health

How?

New comment by rain1 in "Llamafile Not Making Sense"

rain1 — Mon, 22 Apr 2024 08:52:41 +0000

todsacerdoti is a spambot btw

Fixing the volume on my Bluetooth earbuds

rain1 — Sat, 28 Oct 2023 17:06:46 +0000

Article URL: https://blog.ornx.net/post/bluetooth-volume-fix/

Comments URL: https://news.ycombinator.com/item?id=38051488

Points: 293

# Comments: 85

New comment by rain1 in "Why does all() return True if the iterable is empty?"

rain1 — Sat, 26 Aug 2023 13:01:23 +0000

I don't understand this. Please can you point me to information about it?

New comment by rain1 in "Why does all() return True if the iterable is empty?"

rain1 — Sat, 26 Aug 2023 13:00:44 +0000

The people that are astonished by this just need to learn why.

It's not the function that is wrong, it's those people.

New comment by rain1 in "Tricking Monty Hall"

rain1 — Sat, 15 Jul 2023 13:56:51 +0000

This is incorrect, the goats and car are behind doors. They are not inside cardboard boxes.

New comment by rain1 in "Claude 2"

rain1 — Tue, 11 Jul 2023 18:43:48 +0000

This is an example of hallucination.

An LLM doesn't know anything about itself - it can be pre-prompted with facts about itself, but this is going to be an example of it just making plausible text up.