Eva Yi Xie
comp neuro phd student @ princeton. visiting scientist @ Allen Institute. MIT’24. https://minzsiure.github.io
- 🤠 We are in Upper Level Room 10!
- We are over the moon to be joined by so many engaging colleagues across neuro, cog sci, and machine learning, including Terry Sejnowski! #NeurIPS2025
- Reposted by Eva Yi Xie2️⃣ days until #NeurIPS2025 Data on the Brain & Mind workshop! 🧠💭🤖 Join us on Dec 7 for a full-day interactive session 8am-5pm PT. Authors, please remember to RSVP for our mentorship lunch 🥙 generously supported by @kavlifoundation.org and @simonsfoundation.org (@flatironinstitute.org)
- It’s TODAY! 🤗 📍 Exhibit Hall C,D,E #2109 ⏰ 4:30-7:30pm
- Connectome suggests brain’s synaptic weights follow heavy-tailed distributions, yet most analyses of RNNs assume Gaussian connectivity. 🧵⬇️ Our @alleninstitute.org #NeurIPS2025 paper shows heavy-tailed weights can strongly affect dynamics, trade off robustness + attractor dimension.
- 🗓️ Our #NeurIPS2025 poster is in 2 days on Dec 4! “Slow Transition to Low-Dimensional Chaos in Heavy-Tailed Recurrent Neural Networks” 📍 Exhibit Hall C,D,E #2109 ⏰ 4:30-7:30pm We will also be at NeurReps workshop on Dec 7 🙌 Plz come say hi! Happy to chat :) (I’ll also be at UniReps + DBM workshops)
- Connectome suggests brain’s synaptic weights follow heavy-tailed distributions, yet most analyses of RNNs assume Gaussian connectivity. 🧵⬇️ Our @alleninstitute.org #NeurIPS2025 paper shows heavy-tailed weights can strongly affect dynamics, trade off robustness + attractor dimension.
- Connectome suggests brain’s synaptic weights follow heavy-tailed distributions, yet most analyses of RNNs assume Gaussian connectivity. 🧵⬇️ Our @alleninstitute.org #NeurIPS2025 paper shows heavy-tailed weights can strongly affect dynamics, trade off robustness + attractor dimension.
- 1/ Setup: With @mihalas.bsky.social and Lukasz Kusmierz, We study RNNs with weights drawn from biologically plausible Lévy alpha-stable distributions, generalizing the Gaussian distribution to heavy tails.
- 2/ 🔎Result 1: While mean-field theory for the infinite system predicts ubiquitous chaos, our analysis reveals *finite-size* RNNs have a sharp transition between quiescent & chaotic dynamics. We theoretically predict the gain of transition and validated it through simulations.
-
View full thread
