🔥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

Bullshit. Francis won that fight. I don’t care what anyone says. People’s champ. All hail.

Coach of the year Eric Nicksick

She solved the cat's AGI

New OpenAI board just announced! 😍😂

Yann LeCun Geoffrey Hinton It's a shame that it's so hard to get data on this. Maybe a poll? Q1) have you published a paper in a top tier AI venue (NeurIPS/CVPR/ICLR etc). Q2) who do you think is more right about AI safety, Geoff or Yann?

LLMs are so weird because one side is people with five PhDs who have been studying neuron activations for the past three decades and on the other side is someone called leetm5n with an anime avatar just casually releasing increasingly better performing fine tunes of mistral

Thanks for the examiners, and Prof. Noel O'Connor (he/him) for the continuous support. A heartfelt tribute to my supervisor Kevin McGuinness , I deeply wished He was with us today ... His legacy lives on in every aspect of my research, and will always be 🙏

Poirier is a living legend #UFC300

Watch my @stanford CS 25 lecture next week, "aligning open language models," it'll be good, v excited

so cold.

#CVPR2024 DinoV2 is amazing…

BREAKING: 🇷🇸 Novak Djokovic has completed tennis. 1 x Olympics Gold Medal 🥇 and 10 x Australian Open 🇦🇺 3 x Roland-Garros 🇫🇷 7 x Wimbledon 🇬🇧 4 x US Open 🇺🇸 7 x ATP Finals 🏆 8 x year-end No.1s 2 x Career Golden Masters 1 x Davis Cup 🇷🇸 428 weeks at No.1 (most ever)

🚨Don't miss out! Join us on Monday 2nd of September at 13 PM GST for the TII AI seminar with Lucas Beyer (bl16) as he dives into Vision in the age of LLMs, please register here tii.ae/seminar/ai-sem… Technology Innovation Institute

A nice talk full of insights about how vision evolved using language as an “API” from Lucas Beyer (bl16) youtu.be/4NWdzXoEHok

Falcon 3 is out! 1B, 3B, 7B, 10B (Base + Instruct) & 7B Mamba, trained on 14 Trillion tokens and apache 2.0 licensed! 🔥 > 1B-Base surpasses SmolLM2-1.7B and matches gemma-2-2b > 3B-Base outperforms larger models like Llama-3.1-8B and Minitron-4B-Base > 7B-Base is on par with

Here we go 🚀 after months of efforts, we are happy to release the Falcon3 familly. And yes we compare against Qwen Binyuan Hui ✈️ ICLR 2025

Live footage of AI engineers using LLMs to solve every problem... ;)

🚀 Excited to share that our latest paper “Harnessing Frozen Unimodal Encoders for Flexible Multimodal Alignment” has been accepted to CVPR 2025! 🎉 Recent contrastive models like CLIP have set the bar for vision-language understanding. But what if we could achieve similar

It seems like image reps were precomputed if the training follows this github.com/meta-llama/lla… now is this considered as an early fusion? like the moe does not access the images themselves but a stored rep

🚀 Excited to release Freeze-Align code and datasets! Frozen unimodal encoders + lightweight projectors = flexible multimodal alignment. CLIP-level performance 20x less data, 65x less compute! 🔗 Code: github.com/mayug/freeze-a… 🔗 Paper: arxiv.org/abs/2409.19425 #CVPR2025 #AI