ML research scientist @arcinstitute | PhD @PurdueCS | Ex Intern @GoogleDeepMind and @AIatMeta (FAIR)
Mountain View
Joined May 2019
- Tweets32
- Following214
- Followers554
- Likes286
Beatrice Bevilacqua retweeted
Very excited to present our "Differentiable Constraint-Based Causal Discovery" at #NeurIPS2025!
We introduce Differentiable d-Separation, bringing differentiable learning to Constraint-Based Causal Discovery.
Come to chat and learn more!
๐ Hall C,D,E #2407 ๐ Fri 11am-2pm PST
Beatrice Bevilacqua retweeted
Today @arcinstitute releases State, our first perturbation prediction AI model and an important step towards our goal of a virtual cell
State is designed to learn how to shift cells between states (e.g. โdiseasedโ to โhealthyโ) using drugs, cytokines, or genetic perturbations
๐จ New paper at #ICLR2025!
๐งตIntroducing Holographic Node Representations, pretrained general-purpose node embeddings that adapt to any graph task (node, link, high-order)
๐Fri 10โฏAM SGT Hall 3 + 2B #189 presented by @Josh_d_robinson
w/ @Josh_d_robinson @jure @brunofmr
1/8
We also show that HoloGNN can recover canonical eigenvectors (๐ฐ๐ข๐ญ๐ก๐จ๐ฎ๐ญ ๐ฌ๐ข๐ ๐ง/๐๐๐ฌ๐ข๐ฌ ๐๐ฆ๐๐ข๐ ๐ฎ๐ข๐ญ๐ฒ) and use them in downstream tasks, outperforming expensive spectral methods!
7/8
๐ Paper: openreview.net/forum?id=tGYFโฆ
Come by our poster Friday morning #ICLR2025 and chat with @Josh_d_robinson about symmetries and task-agnostic representationsโจ
8/8
If you are at #NeurIPS2024, donโt miss Moshe Eliasof presenting our ๐๐๐๐๐๐๐: ๐๐๐๐ฉ๐ญ๐ข๐ฏ๐ ๐๐จ๐ซ๐ฆ๐๐ฅ๐ข๐ณ๐๐ญ๐ข๐จ๐ง ๐๐จ๐ซ ๐๐ซ๐๐ฉ๐ก ๐๐๐ฎ๐ซ๐๐ฅ ๐๐๐ญ๐ฐ๐จ๐ซ๐ค๐ฌ!
๐
Today 11 am
๐ East Exhibit #3000
Joint work w/ @caromitreka @HaggaiMaron
TLDR: Learn different normalization parameters for every pair of non-isomorphic nodes
How? With an expressive (but scalable!) normalization GNN
Advantages: Adaptivity to the input graph, which results in consistent performance & faster convergence!
Beatrice Bevilacqua retweeted
๐ฃStop by our @NeurIPSConf Poster #3103 Today at 11 AM! for DiGRAF: Diffeomorphic Graph Adaptive Activation Function - an activation function that adapts to the graph structure by learning a diffeomorphism.
Joint Work with Xinzhi Wang (equal contrib.), @caromitreka, @brunofmr , @beabevi_ , and Moshe Eliasof. ๐งต
Do you want to know how to learn to select a ๐ฌ๐ฆ๐๐ฅ๐ฅ ๐ฌ๐ฎ๐๐ฌ๐๐ญ ๐จ๐ ๐ฌ๐ฎ๐๐ ๐ซ๐๐ฉ๐ก๐ฌ for Subgraph GNNs?
Stop by our @iclrconf ๐๐จ๐ฌ๐ญ๐๐ซ #๐๐ ๐๐จ๐๐๐ฒ ๐๐ญ ๐๐:๐๐ to find out!
w/ @brunofmr @HaggaiMaron - arxiv.org/abs/2310.20082
#ICLR2024
Beatrice Bevilacqua retweeted
๐ป๐ธ๏ธToday at #ICML2023, @beabevi_ will present her internship work @GoogleDeepMind; I'll also be there!
We design a causality objective capturing a simple idea: many inputs to a target algorithm will induce identical initial behaviours.
2--3:30pm; Poster #227. See you there! ๐
Don't miss our latest work showing how invariances can once again help us in graph learning!
Now we look at causal tasks!
This Post is from an account that no longer exists. Learn more
๐ง๐ผ๐ฑ๐ฎ๐ ๐ฎ๐ ๐ฐ๐ฃ๐ ๐ถ๐ป ๐๐ฎ๐น๐น ๐ #๐ต๐ฏ๐ต, @ffabffrasca and I will be presenting โUnderstanding and Extending Subgraph GNNs by Rethinking their Symmetriesโ (#NeurIPS2022 Oral)!
Come chat w/ us about Subgraph GNNs, expressivity, and symmetries!
w/ @HaggaiMaron @mmbronstein
Our paper โUnderstanding and Extending Subgraph GNNs by Rethinking Their Symmetriesโ was accepted as an ๐ผ๐ฟ๐ฎ๐น ๐ฝ๐ฟ๐ฒ๐๐ฒ๐ป๐๐ฎ๐๐ถ๐ผ๐ป at #NeurIPS2022!
Check out our recently released code:
github.com/beabevi/SUN
arxiv.org/abs/2206.11140
Overwhelmed by a plethora of similar, expressive GNN architectures? Symmetries and equivariance to the rescue again! ๐งต
A fresh, unifying perspective on Subgraph GNNs appearing at #NeurIPS2022 ๐
Joint work with: @beabevi_ * @mmbronstein @HaggaiMaron โ arxiv.org/abs/2206.11140
