🔥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

📊 See the figure to learn how our GOKU-UI beats the baselines in reconstruction while identifying the governing dynamics. A step towards interpretable forecast:

Interview in nature on how we are developing precision child psychiatry at Centre de recherche Azrieli du CHU Sainte-Justine/Université de Montréal. My colleagues Dr.Conrod speaks on imaging, Sebastien Jacquemont on genetics and I discuss Precision Psychiatry & Social Physiology lab. work on computational psychiatry & hyperscanning for the clinic nature.com/articles/d4247…

Neural spiking data and Transformers are a tricky match. Temporal segmentation and tokenization are the crux. Together with an all-star team, we figured out a scalable sol. The results are exciting: training and transferring on multi-sessions, multi-subjects neural decoding tasks

I'm not at #NeurIPS2023 in-person this year, but Germán Abrevaya is, and he'll present our paper as a spotlight talk in a few hours👀. He was @ ML and Physics workshop today and is your expert for all the latent differential equation ideas. Your final chance to connect with him👇

Thrilled to announce our paper on integrating domain knowledge into machine learning models & encoding whole-brain dynamics into low-dimensional dynamical systems is now published at Transactions on Machine Learning Research! 🎉 openreview.net/forum?id=uxNfN… youtu.be/XYv10fuumCQ #SciML #TimeSeries #Neuroscience

🚨 Open-source AI community - stop building everything from scratch, let's build on each other's (and your own) work over time - continually, as we should! Don't waste previoius compute and human effort! See arxiv.org/abs/2403.08763 - simple & useful tips on how to just keep

Very excited to be at ICLR 2026 and AISTATS Conference in the next 2 weeks, presenting 3 papers - 2 on causal representation learning, and one on compositional+length generalization of seq2seq models! Looking forward to reconnecting with many old friends. Come check our work! 1/5

Introducing TACTiS-2: Better, Faster, Simpler Attentional Copulas for Multivariate Time Series. A highly flexible model for multivariate probabilistic time series prediction. To be presented at ICLR 2024. Find us at Hall B #145 on Thu 9 May 10:45 a.m. - 12:45 p.m. Details🧵👇

Distinct social behavior and inter-brain connectivity in dyads with autistic individuals tandfonline.com/doi/full/10.10… 🍾 New @PPSP_Team paper out! — 12 years in the making... thread!🧵⬇️

That was a blast! Loved the mind jazz, and discovering many new lines of research from fellow scientists. Thank you so much for the invitation 🙏

Just out: “Non-invasive #EEG in awake #cats” sciencedirect.com/science/articl… — The first base is secured to move to inter-species #hyperscanning recordings! :) #DogsNext #StayTuned

First lab meeting of the Academic year! Kudos to Anne Monnier & Camille Lanoix for their stimulating talk on Generative #Neurophenomenology and how we study it at @PPSP_Team! Roadmap paper submitted… #StayTuned #NowSTheTime

Building on other people's novel machine learning algo.s is like walking on a minefield. I explained this in a UNIQUE Center Neuro-AI symposium, "From Machines to Brains". The problem is inherent to the requirements for publishing in ML fields. arxiv.org/abs/2407.12220

Brains as dynamical systems! 💯

How does #AI compare to the nuanced reasoning of clinicians in #Psychiatry? Our latest #Review revisits Meehl’s factors, bridging the human and machine reasoning gap to advance interdisciplinary collaboration doi.org/10.1016/j.psyc…🧠🤖 Big kudos to Gauld Christophe & dream team!

See you on the other side💙. mahtaao.bsky.social

How can we design human-interpretable AI models to explain why neurons behave as they do? Excited to share our recent work, a collaboration between three research groups from Harvard University. Sara Matias Hao Wu Simona Temereanca @naoshigeuchida MurthyLab at Harvard Paul Masset (Dem + 1) x Ba 1/21

Introducing nanoAhaMoment: Karpathy-style, single file RL for LLM library (<700 lines) - super hackable - no TRL / Verl, no abstraction💆‍♂️ - Single GPU, full param tuning, 3B LLM - Efficient (R1-zero countdown < 10h) comes with a from-scratch, fully spelled out YT video [1/n]

🧵 1/ How well do LLMs actually do on Olympiad-level math? We evaluated frontier models on 455 problems from the IMO Shortlist. Unlike most benchmarks, we emphasize proof validity, not just final answer correctness. Here’s what we found 👇

I became Canadian thanks to University of Waterloo “re-opening” post-Trump in 2017. Hope other Canadian institutions follow suit in hauling talent once again. NRC Canada ISED