🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Looking forward to presenting Tree-WSV for fast unsupervised ground metric learning at ICLR 2026 2025 tomorrow at 3 pm! Come find me at poster #197 🌳

📍 I’ll be presenting our AISTATS 2025 paper “Credal Two-sample Tests of Epistemic Uncertainty” this week in Phuket! Joint work with the amazing Antonin Schrab , Arthur Gretton , Dino Sejdinovic , and JUNJIRA MUANDET 🙌 paper: arxiv.org/pdf/2410.12921 video: youtu.be/Rq9qW0GZJeE

Density Ratio-based Proxy Causal Learning Without Density Ratios 🤔 at #AISTATS2025 An alternative bridge function for proxy causal learning with hidden confounders. arxiv.org/abs/2503.08371 Bariscan Bozkurt, Ben Deaner, Dimitri Meunier LY9988

📢 We have an opportunity for students to join our PhD programme in Theoretical Neuroscience and Machine Learning this September. Application deadline is 27 May 2025. Information & how to apply 👉 ucl.ac.uk/gatsby/study-a…

Credal Two-Sample Tests of Epistemic Uncertainty at #AISTATS25 Compare credal sets: convex sets of prob measures where elements capture aleatoric uncertainty and set itself represents epistemic uncertainty. arxiv.org/abs/2410.12921 Siu Lun Chau Antonin Schrab Dino Sejdinovic Krikamol (Hiring Postdoc)

Kernel Single Proxy Control for Deterministic Confounding at #AISTATS2025 Proxy causal learning generally requires two proxy variables - a treatment and an outcome proxy. When is it possible to use just one? arxiv.org/abs/2308.04585 LY9988

Oral #AISTATS25 Robust Kernel Hypothesis Testing under Data Corruption -Robustify any permutation test to be immune to data corruption of up to X samples -MMD HSIC robust minimax optimality Monday 5 May -Oral Session 7 Robust Learning -Poster Session 3 Presented by Ilmun Kim

Our paper "Solve Layerwise Linear Models First to Understand Neural Dynamical Phenomena" has been accepted as a position paper to ICML 2025! arxiv.org/abs/2502.21009 These models offer a tractable path to understanding complex neural dynamics—before diving into full nonlinearity.

SWC researchers have uncovered a new way that the brain learns. The second learning system is frequency-based and may help explain how the brain forms habits and why they are so hard to break. Full story ⤵️ sainsburywellcome.org/web/research-n…

Our paper just came out in PRX! Congrats to Nishil Patel and the rest of the team. TL;DR : We analyse neural network learning through the lens of statistical physics, revealing distinct scaling regimes with sharp transitions. 🔗 journals.aps.org/prx/abstract/1…

Square loss, heavy-tailed noise... not as bad as you think!

🚨 New paper accepted at SIMODS! 🚨 “Nonlinear Meta-learning Can Guarantee Faster Rates” arxiv.org/abs/2307.10870 When does meta learning work? Spoiler: generalise to new tasks by overfitting on your training tasks! Here is why: 🧵👇

How does in-context learning emerge in attention models during gradient descent training? Sharing our new Spotlight paper ICML Conference: Training Dynamics of In-Context Learning in Linear Attention arxiv.org/abs/2501.16265 Led by Yedi Zhang with Aaditya Singh and Peter Latham

Composite Goodness-of-fit Tests with Kernels, now out in JMLR! jmlr.org/papers/v26/24-… Test if your distribution comes from ✨any✨ member of a parametric family. Comes in MMD and KSD flavours, and with code. Oscar Key François-Xavier Briol Tamara Fernandez

Missing ICML due to visa :'(, but looking forward to share our ICML paper (arxiv.org/abs/2502.05318) as a poster at #BayesComp, Singapore! Work on symmetrising neural nets for schrodinger equation in crystals, with the amazing Zhan Ni, Elif Ertekin, Peter Orbanz and Ryan Adams

Meanwhile, excited to be in #Lyon for #COLT2025, with a co-first author paper (arxiv.org/abs/2502.15752) with the amazing team -- Matthew Mallory and our advisor Morgane Austern! Keywords: Gaussian universality, dependent data, convex Gaussian min-max theorem, data augmentation!

Last but not least of my travel updates: Courtesy of the very kind Siu Lun Chau, I'm giving a talk at 14:30, 27 Jun NTU College of Computing and Data Science (CCDS) in SG on data augmentation & Gaussian universality, which strings together several works over my PhD. If you're in SG/Lyon the next few weeks, let me know!

Excited to share this work has been accepted as an Oral at #icml2025 -- looking forward to seeing everyone in Vancouver, and an extra thanks to my amazing collaborators for making this project so much fun to work on :)

New research shows long-term learning is shaped by dopamine signals that act as partial reward prediction errors. The study in mice reveals how early behavioural biases predict individual learning trajectories. Find out more ⬇️ sainsburywellcome.org/web/blog/long-…