I think Claude 3 crossed (or closely approached) an interesting threshold:

The "power users" threshold.

For the first time being able to help power users with heavy complex tasks faster than themselves.

This is a controversial topic in AI,
Let me try to explain: I never use GPT-4 for code.

I use it for:
1. Brainstorming ideas.
2. Learning new topics that I don't know about.
3. Reading long texts instead of me (ask-your-pdf)
4. Just easy tasks.

But never for code.
It never helped me coding.

Gazans going crazy when realizing there are Muslims in the IDF

A Thread 🧵👇 "Are these Muslims like us or what is their religion exactly?"

First week in the army.

Some thoughts:

The IDF is unbelievably different now than what is was during my (5yr) service.

Unrecognisable.

No clutter.

No bloat.

No BS procedures.

Everything (and everyone) is insanely efficient and highly motivated. Young soldiers are not complaining and oftentimes also overdo their roles voluntarily.

More reservists keeps showing up.

(Many of which volunteered and were not supposed to attend otherwise)

Civilians show up every couple of hours with food, clothing and supplies.

The strongest model you didn't know about:

- The best model.
- Trained on the best dataset.

OpenChat V2 x OpenOrca

OpenChat V2, A leading open-source model was fine-tuned last night on Open-Orca's data.

Link: https://t.co/4hiMTqXUOJ https://t.co/jG9Abqeos0huggingface.co/openchat/openc…
twitter.com/i/web/status/1…
OpenChat?

A series of powerful open source models leading multiple leader boards.

The teams' latest model, OpenChat V2 (weighted), Is currently leading 4 different benchmarks.

Github:
Model: https://t.co/Fv0793EEQSgithub.com/imoneoi/opench…
huggingface.co/openchat/openc…

Reverse engineering Code-Interpreter

Last night I participated in an interesting twitter space: Reverse Engineering the code-interpreter

Here are some of the most interesting bits: https://t.co/pUudeV5uuXtwitter.com/i/web/status/1…
Since the new model is able to run code, many people have already managed to ask it nicely for all the information it has inside the virtual machine it runs on.

I think I get it now. let's try something out:
.
Comment to this thread everything you "just guess" about GPT-4 👇👇 Guess: In the dataset: Went through every undergrad major and included exams, tests & textbooks of the subject. [to create a "wow" effect for every educated person no matter the field]

Ladies and Gentlemen. GPT-4 👇 *Ported from HF's code

B R E A K I N G! @realGeorgeHotz Just said:
- GPT-4 Has 220 parameters.
- GPT-4 Is a mixture of experts with 8 experts.
- GPT-4 Is doing 16 times inference (did u mean beams? or just 2 beams per model?)
-
@realGeorgeHotz It's HUGE!! Can you confirm??
src: latent.space/p/geohot twitter.com/i/web/status/1… @abacaj look what you have done! I had work to do!

החלקים החסרים של GPT-4 (בעברית) 👇
.
The missing pieces of GPT-4 (Hebrew)👇
.
​טריקים שאף אחד לא מדבר עליהם ועתיד מודלי השפה הפתוחים.
***
> אמ;לק: למידה ממשוב לא-אנושי (!!)
.
העברתי הרצאה מעניינת לפני כמה ימים בכנס #GenML עם אותו השם ("החלקים החסרים של GPT-4").
בזמן שכולנו מחכים… twitter.com/i/web/status/1… חלק 2: ההשפעה של LLaMA
----------
אתם כבר יודעים על LLaMA.
מודל שפה עוצמתי מבית מטא, שוחרר בקוד פתוח בפברואר האחרון.
המודל עומד במרכזו של מאמץ עולמי לשחזר את יכולות מודלי השפה המסחריים באופן פתוח.
בכך לספק למיליוני אנשים בכל רחבי העולם גישה לטכנולוגיה זו.
ישנם עשרות אלפי אנשים… twitter.com/i/web/status/1…

How to trick AUC into scoring your stupid models higher 👌

👇A - Thread - about gaming the AUC metric!👇

There is a built-in "flaw" with the AUC metric. Today we are going to take it for a spin! One of the questions coming up as we get stuck on improving individual models further is how to best combine multiple models.

Apart from simple the baseline Avg, Median or Avg-Median-Rocket-Science-Blending[1].