Adam Perold @AdamPerold

Joined June 2012

Tweets

127
Followers

301
Following

1K
Likes

8K

Adam Perold @AdamPerold

2 months ago

This model and system prompt @tobi 🌟😆

I love this idea from Tobi Lutke: “I take the view that I'm a corporate raider. That Shopify went bankrupt and I bought it on a fire sale and I'm marching in on day one, and that previous management was crazy and we need to turn this place around.”

26 88 2K 198K 1K

0 0 0 158 0

View Details

Adam Perold @AdamPerold

3 months ago

@WillSmithVision @sigbovik 😂🤣

0 0 0 333 0

View Details

Adam Perold @AdamPerold

4 months ago

A new pathway to intelligent systems. This is the way. Congrats @ylecun, @sainingxie and team.

AMI Labs @amilabs

4 months ago

Advanced Machine Intelligence (AMI) is building a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe. We’ve raised a $1.03B (~€890M) round from global investors who believe in our vision of universally

340 874 8K 5.0M 3K

0 0 2 257 0

View Details

Adam Perold @AdamPerold

4 months ago

Inspiring team ✨🌲

Robin Rombach @robrombach

4 months ago

New paper out! We present a training method for multimodal generative models, called Self-Flow, which combines classic flow matching and representation learning. Why? Unlike most representation alignment methods, our new approach does not require external, pretrained models and

5 36 309 28K 145

1 0 3 301 0

View Details

Adam Perold @AdamPerold

5 months ago

@miramurati Congrats @soumithchintala !! 💥

0 0 0 286 0

View Details

Pascale Fung @pascalefung

6 months ago

Introducing VL-JEPA: Vision-Language Joint Embedding Predictive Architecture for streaming, live action recognition, retrieval, VQA, and classification tasks with better performance and higher efficiency than large VLMs. • VL-JEPA is the first non-generative model that can perform general-domain vision-language tasks in real-time, built on a joint embedding predictive architecture. • We demonstrate in controlled experiments that VL-JEPA, trained with latent space embedding prediction, outperforms VLMs that rely on data space token prediction. • We show that VL-JEPA delivers significant efficiency gains over VLMs for online video streaming applications, thanks to its non-autoregressive design and native support for selective decoding. • We highlight that our VL-JEPA model, with an unified model architecture, can effectively handle a wide range of classification, retrieval, and VQA tasks at the same time. by @Delong0_0 @MustafaShukor1 @TheoMoutakanni @willyhcchung Jade Lei Yu Tejaswi Kasarla @AllenBolourchi @ylecun @pascalefung arxiv.org/abs/2512.10942

10 85 553 90K 394

View Details

Adam Perold @AdamPerold

7 months ago

@qasar 💥💥

0 0 0 151 0

View Details

General Intuition @gen_intuition

7 months ago

Yann LeCun (Chief AI Scientist, Meta, @ylecun), @PimDeWitte (CEO, General Intuition), and Aude Durand (Kyutai, @aude_drn), talk about world models, embodied agents, Yann's new company, and the limitations of LLMs 0:00 - Introduction to World Models 5:00 - Why World Models, Intuition & Introducing Yann's new company 10:00 - Architectures + Merging Language & Interaction Data towards General Agents 20:00 - Open Source, Sovereign AI & @kyutai_labs Partnership Keynote for #aiPULSE2025 at Station F in Paris 🇫🇷

29 123 642 111K 607

View Details

Adam Perold @AdamPerold

7 months ago

@robrombach 🌲🌟🌟🌟

0 0 1 35 0

View Details

Adam Perold @AdamPerold

7 months ago

@andi_blatt 🌲🌟🌟🌟

0 0 1 19 0

View Details

Adam Perold @AdamPerold

7 months ago

@soumithchintala Congrats!

0 0 0 126 0

View Details

Adam Perold @AdamPerold

7 months ago

@soumithchintala @deedydas @psermanet 💛

0 0 0 1K 0

View Details

Adam Perold @AdamPerold

9 months ago

Great decomposition @karpathy

Andrej Karpathy @karpathy

9 months ago

Finally had a chance to listen through this pod with Sutton, which was interesting and amusing. As background, Sutton's "The Bitter Lesson" has become a bit of biblical text in frontier LLM circles. Researchers routinely talk about and ask whether this or that approach or idea

415 1K 9K 2.0M 9K

0 0 0 419 0

View Details

Adam Perold @AdamPerold

9 months ago

@nicochristie @satyanadella 💥💥

0 0 0 162 0

View Details

Adam Perold @AdamPerold

9 months ago

Congrats @robrombach @andi_blatt ! 🌲💥

Robin Rombach @robrombach

9 months ago

FLUX is now also in Meta AI. What a day. 🥵

16 8 185 28K 21

0 0 3 501 0

View Details

Adam Perold @AdamPerold

a year ago

@Andersonmancini @levinriegner @TheWebbyAwards Congrats! 🌟💫

1 0 1 42 0

View Details

Adam Perold @AdamPerold

a year ago

There’s a nice feature in Chrome, where you just right click on the tab and select “Add tab to group,” then name the group, which you can do with a symbol like “+” or “.” for simplicity. A clever design is that you only need to do this twice. Once you have two tabs in a group displaying, any other tabs you drag between them automatically assign to the group. So you can group-select and assign 200+ tabs pretty effortlessly. I assume this wouldn’t recover tabs from a “remove browsing profile” action, which sounds like a privacy feature intended to fulfill the result you achieved. But I thought this was a nice solution to the tabs issue worth sharing.