@RichDecibels Or, it shows how they think about society? They only care about people they know, we care about all individuals that are part of our society
Modern deep networks are often trained at the #EdgeOfStability, a regime where dynamics are locally unstable, nearing chaos. Yet generalization improves, defying the wisdom of classical optimization. We now theoretically explain this central puzzle: https://t.co/yh79EN8EfF. 👇
I ran the numbers on NeurIPS 2025 myself. 15.4% of papers (292/1892) had at least one Iranian co-author. Iranians are just 0.9% of US immigrants. That's the highest ratio of any immigrant group. Sample of papers attached
It is honestly alarming to me that stuff like this, the idea that we ought to significantly delay curing cancer exclusively to give human researchers the personal gratification of finding it without AI, is being taken seriously at conferences
1/2) It's finally out on Arxiv: Feedback guidance of generative diffusion models!
We derived an adaptive guidance methods from first principles that regulate the amount of guidance based on its current state.
Complex prompts are highly guided while simplem ones are almost free
Why academia is sleepwalking into self-destruction. My editorial @Brain1878 If you agree with the sentiments please RT. It's important for all our sakes to stop the madness
https://t.co/5S5rL8Yda1
Our new topological, diffusion-bridge based graph generation framework, HOG-Diff [https://t.co/gOGvlA7F5z], leverages higher-order information as a guide to progressively generate plausible graphs thru a coarse-to-fine curriculum. #TDL#TopologicalDeepLearning#GenAI#ML#AI 👇
@canaesseth Very interesting work, congrats!
Question: you need to learn a function F_\phi(e, t, X) that gives you the latent z_t directly? (eq. 16)
Would it be fair to say you 'learn' to solve SDEs implicitly, inside this function?
I'am thrilled to announce that our paper "Generative Fractional Diffusion Models (GFDM)" has been accepted at #NeurIPS2024! 🚀
📜Check-out our paper: https://t.co/0p4alP3dRj
🌟Star our repository for code release: https://t.co/pRHdZL9KyQ
🧵1/6
@ManuelSintubin Helaas merk je dat enkel op als je zelf iets kent van het onderwerp, maar in hoeverre is dat zo voor al die zaken waar je zelf niets over weet? 🤔
Excited to present our paper, "Generative Fractional Diffusion Models (GFDM)" [https://t.co/a0VA70UjTE], at the @icmlconf Workshop SPIGM. Join us at our poster session on Friday from 11:30 to 12:30 in Lehar 3 or reach out directly to me to catch up at ICML on another occasion.
I will be at @icmlconf next week presenting 2 papers and 2 workshop papers. I'm also on the job market! Please reach out if you're interested in discussing generative models, geometric DL, or AI4Science.
What I find slightly depressing is that none of what is written here is *complicated* and many 7yo could grasp this.
There is a strange cultural systemic issue that leads us as a society to expose most of the kids to 10% of what they could actually understand in math.