iiinova — where your digital self keeps evolving

Jacob Bamberger

@jacobbam.bsky.social

371 followers 143 following 11 posts

ML PhD student at University of Oxford. Interested in Geometric Deep Learning

Posts Replies Media

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

Presenting at poster session 4 east. 📅Wednesday, July 16th 🕓4:30-7:00 PM 📈#E-2802

000

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

Read more here: 📄paper: arxiv.org/abs/2506.05971 💻 code: github.com/BenGutteridge/… 🙌 With Ben Gutteridge, Scott le Roux, @mmbronstein.bsky.social, Xiaowen Dong #ICML2025 #GNN #AI

arxiv.org

On Measuring Long-Range Interactions in Graph Neural Networks

Long-range graph tasks -- those dependent on interactions between distant nodes -- are an open problem in graph neural network research. Real-world benchmark tasks, especially the Long Range Graph Ben...

122

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

🔑 Takeaways: ✅ Long-range can be formalized & measured ✅ Reveals new insights into models & datasets 🚀 Time to rethink evaluation: not just accuracy, but how models solve tasks

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

Why does this matter? "Long-range" is often just a dataset intuition or model label. We offer a measurable way to: 💡Understand models 🧪Test benchmarks 🦮Guide model design 🚀Go beyond performance gaps

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

We reassess LRGB, the go-to long-range benchmark, by checking if model range correlates with performance—expected for truly long-range tasks. Surprisingly: ❌ Peptides-func: negative correlation, suggests not long-range ✅ VOC: positive correlation, suggests long-range

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

We validate our framework in three steps: 👷Construct synthetic tasks with analytically-known range 💯Show trained GNNs can approximate the true task range 🔬Use range as a proxy to evaluate real benchmarks

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

Our measure uses the model's Jacobian (for node tasks) and Hessian (for graph tasks) to quantify input-output influence, works with any distance metric, and supports analysis at all granularities—node, graph, and dataset.

110

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

We propose a formal range measure for any graph operator, derived from natural axioms (like locality, additivity, homogeneity) — and show it’s the unique measure satisfying these. This measure applies to both node- and graph-level tasks, and across architectures.

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

"Long-range tasks" are a central yet vague challenge in graph learning. What makes a task long-range? How can we tell if a model actually captures long-range interactions?

100

Jacob Bamberger @jacobbam.bsky.social · 13/07/2025

🚨 ICML 2025 Paper 🚨 "On Measuring Long-Range Interactions in Graph Neural Networks" We formalize the long-range problem in GNNs: 💡Derive a principled range measure 🔧 Tools to assess models & benchmarks 🔬Critically assess LRGB 🧵 Thread below 👇

150

Reposted by Jacob Bamberger

Fabrizio Frasca @ffabffrasca.bsky.social · 03/12/2024

🌟 GLOW is coming back in December with amazing speakers: Emily Jin and @joshsouthern.bsky.social ! 🗓️ Dec 18th @ 17 CET on Zoom, don't miss that! 🌐 Find more here: sites.google.com/view/graph-l...

1144

Jacob Bamberger @jacobbam.bsky.social · 25/11/2024

It’s a wrap! Thank you to everyone that joined LoG-ox, the Oxford local meet up for @logconference.bsky.social!

1202