🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Can't wait for your feedback on our largest and most capable Phi model yet! huggingface.co/microsoft/Phi-…

Very strong results on MMLU-Pro! It's even higher than 70B models.

Our MMLU-Pro paper is out. It's a more difficult, robust and reasoning-driven benchmark to measure expert-level intelligence. We have gradually included 50+ models in our leaderboard: huggingface.co/spaces/TIGER-L…. GPT-4o, Gemini-1.5-Pro, Claude-3-Opus are the current top-3 models.

👉 Structured generation beats GPT-4 Using structured generation, phi-3 achieves 95.5% accuracy when it only achieves 86% without structured generation. More importantly, it beats GPT-4 (93.5%) by a whopping 2 percentage point. 🔥🔥🔥

Every day I witness the AI revolution in action, and every day I see 1 or 2 questions that would deserve an entire PhD thesis to explore fully ... Honestly, how lucky we are to do research in that era!! *Even if* there is no more magical leap like from gpt2 to 3, or 3 to 4,

Just to state the obvious, AGI might not happen with current techniques. At a very basic level the argument for AGI is: if you compress sufficiently complex material, like the web, then you end up creating a "mind" that has in it all the "operations" necessary to create the web,

Claude 3 Opus vs 3.5 Sonnet: tikz draw Golden Gate

Some of you might find this short interview interesting. Quite a journey with Ronen Eldan since the entropic barrier 🤣, and it's still just the beginning! turingpost.com/p/phi3

It took more than a year but maybe Sparks experiments are finally reproducible 😁

If you're attending ICML 2024, join my 2-hour tutorial on Monday July 22 to explore the Physics of Language Model - all 6 parts. Visit: physics.allen-zhu.com and it will be live-streamed on Zoom. BONUS: this is the premiere of Part 2.1 + 2.2, don't miss out! #ICML2024 #MetaAI

Second Ironman 70.3 in the bag 😁. Quite happy to have finished this one, it demanded some sacrifices during a busy year in AI... To the youngsters out there, don't forget to get your miles in even when work gets busy 😅

The latest update to phi-3-mini is worth checking out! Better instruction following, more robust, etc...

Check out this podcast about SLMs, it's a bit more product/business oriented than what I usually talk about but hopefully should be interesting to some of you! podcasts.apple.com/us/podcast/mic…

amathr.org/prizes/aiprize/ The Association for Mathematic Research announces "Prize in the Mathematics of Artificial Intelligence". I'm in the selection committee. The goal is to inspire young people to work on the intersection of AI and maths. Nominations to [email protected]

We released phi 3.5: mini+MoE+vision A better mini model with multilingual support: huggingface.co/microsoft/Phi-… A new MoE model:huggingface.co/microsoft/Phi-… A new vision model supporting multiple images: huggingface.co/microsoft/Phi-…

This is the coolest "application" of LLMs I have seen in a while: instead of using an LLM to solve a problem set, flip the table and ask the student to TEACH the LLM how to solve the problem set. Really lovely. Thx to Jordan Hoffmann for sharing this :-). youtube.com/watch?v=om7VpI…

Lots of progress on bandit convex optimization recently arxiv.org/abs/2406.18672 arxiv.org/abs/2406.06506 arxiv.org/abs/2302.05371, I wish I could follow it more closely ... looks like Conjecture 1 from arxiv.org/abs/1607.03084 is going to be resolved soon!!!