Gennaro Gala
@gengala13
AI researcher https://gengala.github.io/
I'm constantly irritated that I don't have time to read the torrent of cool papers coming faster and faster from amazing people in relevant fields. Other scientists have the same issue and have no time to read most of my lengthy conceptual papers either. So whom are we writing…
I see a paper that looks interesting I download the paper Open the link to the code "Code coming soon" one year ago Delete the paper Memorize the authors' names to avoid wasting time in the future
Nobody wants to hear it, but working on data is more impactful than working on methods or architectures.
1. We often observe power laws between loss and compute: loss = a * flops ^ b + c 2. Models are rapidly becoming more efficient, i.e. use less compute to reach the same loss But: which innovations actually change the exponent in the power law (b) vs change only the constant (a)?
We propose Neurosymbolic Diffusion Models! We find diffusion is especially compelling for neurosymbolic approaches, combining powerful multimodal understanding with symbolic reasoning 🚀 Read more 👇
feeling a bit under the weather this week … thus an increased level of activity on social media and blog: kyunghyuncho.me/i-sensed-anxie…
Scaling continuous latent variable models as probabilistic integral circuits @tetraduzione @gengala13
Less than two weeks to submit your papers on: 📈 #lowrank adapters and #factorizations 🧊 #tensor networks 🔌 probabilistic #circuits 🎓 #theory of factorizations to the first workshop on connecting them in #AI #ML at @RealAAAI please share! 🔁 👇👇👇 april-tools.github.io/colorai/
many of the recent successes in #AI #ML are due to #structured low-rank representations! but...What's the connection between #lowrank adapters, #tensor networks, #polynomials and #circuits? join our @RealAAAI workshop to know the answer! 👇👇👇 april-tools.github.io/colorai/
factorizing a tensor with infinite dimensions⁉️ yes that's possible‼️ and this gives you tractable inference, and the simplest and most principled way to train probabilistic circuits at scale✅ spotlight at #NeurIPS24 📜openreview.net/forum?id=Ke40k… 🖥️github.com/gengala/ten-pi…
I’ll be attending #NeurIPS2024, where I’ll present our spotlight: Scaling Continuous Latent Variable Models as Probabilistic Integral Circuits (PICs). TL;DR: We learn continuous hierarchical mixtures as DAG-shaped PICs, and scale them using neural functional sharing techniques.
Lorenzo & team are really pushing the boundaries of tractable models, check out this banger!
We learn more expressive mixture models that can subtract probability density by squaring them 🚨We show squaring can reduce expressiveness To tackle this we build sum of squares circuits🆘 🚀We explain why complex parameters help, and show an expressiveness hierarchy around🆘