Aleksandar Bojchevski

1/5 Ever wondered how to apply conformal prediction when there's epistemic uncertainty? Our new paper addresses this question!CP can benefit from models like Bayesian, evidential, and credal predictors to have better prediction sets,for instance, in terms of conditional coverage.

🚨 Robust conformal prediction is expensive as we need around 10000 forward passes per input. Or Is it? Checkout our ICLR2025 paper: openreview.net/forum?id=ltrxR… We extend conformal sets to worst case noise under any smoothing, with much less samples. Joint work with @abojchevski

Want to learn continuous & discrete Flow Matching? We've just released: 📙 A guide covering Flow Matching basics & advanced methods arxiv.org/abs/2412.06264. 💻 An open source codebase with image & text examples github.com/facebookresear…. 🗣️ A Flow Matching tutorial #NeurIPS2024.

Missed out on #Swift tickets? No worries—swing by our #SVFT poster at #NeurIPS2024 and catch *real* headliners! 🎤💃🕺 📌Where: East Exhibit Hall A-C #2207, Poster Session 4 East ⏲️When: Thu 12 Dec, 4:30 PM - 7:30 PM PST #AI #MachineLearning #PEFT #NeurIPS24

A common question nowadays: Which is better, diffusion or flow matching? 🤔 Our answer: They’re two sides of the same coin. We wrote a blog post to show how diffusion models and Gaussian flow matching are equivalent. That’s great: It means you can use them interchangeably.

@guennemann kicking of the LOG meetup in Aachen.

@mundt_martin Congratulations!

🎉 Exciting major personal news 🎉 I’ll be joining the University of Bremen as a full (tenured) professor for Lifelong Machine Learning in January 25! The focus will be on researching adaptive, sustainable & inclusive AI. -> apps for 2 new PhD students will open next week!

1/ Can Large Language Models (LLMs) truly reason? Or are they just sophisticated pattern matchers? In our latest preprint, we explore this key question through a large-scale study of both open-source like Llama, Phi, Gemma, and Mistral and leading closed models, including the recent OpenAI GPT-4o and o1-series. arxiv.org/pdf/2410.05229 Work done with @i_mirzadeh, @KeivanAlizadeh2, Hooman Shahrokhi, Samy Bengio, @OncelTuzel. #LLM #Reasoning #Mathematics #AGI #Research #Apple

Move over Taylor there's another SVFT in town.

I am happy to announce that @abojchevski, Michael Schaub, Martin Grohe, and I are organizing a @LogConference local meet-up in Aachen (27.11.-28.11.), including keynotes from @guennemann and @BurkholzRebekka. See log-rwth.github.io/log-meetup-202… for more information.

✨🎨🏰Super excited to share our new paper Ensemble everything everywhere: Multi-scale aggregation for adversarial robustness Inspired by biology we 1) get adversarial robustness + interpretability for free, 2) turn classifiers into generators & 3) design attacks on vLLMs 1/12

#ICML2024 With conformal prediction (CP) we return sets that include the true class with guaranteed high probability, With robust CP we maintain this probability even for worst case (adversarial) noisy input. With CAS (our method) we have robust CP with efficient (small) sets.

🎉Happy to announce that our paper – SVFT: Parameter-Efficient Fine-Tuning with Singular Vectors has been accepted for an oral presentation at WANT @icmlconf!✨#ICML2024 #MachineLearning Paper: openreview.net/forum?id=DOUsk…

🤔 Ever wondered how vulnerable Graph Neural Networks (GNNs) really are? 🚀 Check out our slightly late, but crucial, #ICLR2024 paper: "Rethinking Label Poisoning for GNNs: Pitfalls and Attacks"! 🔥

Read all about our highly parameter-efficient fine tuning approach that Pareto-dominates LoRA and other baselines.

This paper made me smile a lot while working on it, so I want to share a bit about it arxiv.org/abs/2405.09676. We draw a parallel story to the Eckart-Young Theorem (from numerical analysis) in stochastic optimization/learning problems. (with Josh Cutler and Dima Drusvyatskiy)

Another entry in a long-running series where Nicholas Carlini breaks ML defenses published at top security conferences with as little effort as possible (in this case a one line bugfix in the eval)

We introduce Spatio-Spectral GNNs (S²GNNs) – an effective modeling paradigm via the synergy of spatially and spectrally parametrized graph conv. S²GNNs generalize the spatial + FFT conv. of State Space Models like H3/Hyena. Joint work w/ @ArthurK48147 @dan1elherbst @guennemann

On tasks like coding we can keep increasing accuracy by indefinitely increasing inference compute, so leaderboards are meaningless. The HumanEval accuracy-cost Pareto curve is entirely zero-shot models + our dead simple baseline agents. New research w @sayashk @benediktstroebl 🧵

Excited to announce a new paper in PNAS today! An illusion of predictability in scientific results: Even experts confuse inferential uncertainty and outcome variability pnas.org/doi/10.1073/pn… Joint w/ great coauthors @_szhang, @P_Heck, @MichelleNMeyer, @cfchabris & @dggoldst