@bendfulcher.bsky.social
I lead the Dynamics and Neural Systems Group at the School of Physics, the University of Sydney. We develop time series tools & physical models to understand the dynamics of complex (usually neural) systems. Also: @[email protected]
@estherschindler.bsky.socialReposted by @bendfulcher.bsky.socialRIP. Margaret Hamilton, computing pioneer who led software development for the Apollo program, dies at 90
@loopyluppi.bsky.socialReposted by @bendfulcher.bsky.social1/ 🚨 Just out in @natneuro.nature.com! 🧠
We profiled >6,000 features of brain dynamics across humans, macaques, marmosets, mice, fish and worms 🧑🐒🐒🐁🐟🪱 to find a shared neural endpoint of anaesthesia: spatiotemporal isolation of local activity 🧵👇
@bendfulcher.bsky.socialIt’s hard to capture the diversity of complex dynamical patterns in just 2 dimensions but this is my best attempt!
@goliashf.bsky.socialReposted by @bendfulcher.bsky.socialSo excited that our work is now published in PLOS Biology!! 🎉
Check out the final version here: doi.org/10.1371/jour...
And for a bit more context and what we found, here’s our earlier thread: bsky.app/profile/goli...
@bendfulcher.bsky.socialHaha good luck on the quiz! 😅
@epimechfc.bsky.socialReposted by @bendfulcher.bsky.socialCan mathematics explain how biological patterns are formed? Alan Turing thought so. Join me, @benswedlund.bsky.social, as we explore his reaction-diffusion theory and how it transformed our understanding of pattern formation and self-organisation.
@bendfulcher.bsky.socialNow published in PLoS Computational Biology!
A cortical model combining geometric wave propagation with connectome-wired fast-conducting non-local projections (which strongly shape rapid stimulus-evoked responses but contribute little to slower spontaneous fluctuations)
@taramurphy.bsky.socialReposted by @bendfulcher.bsky.socialWe are offering a new round of 4 Physics Foundation PhD scholarships @sydney.edu.au
Available in any area of #physics or #astronomy. Please send to any students (domestic and international) who may be interested.
Deadline 18th October.
⚛️ ☄️
@bendfulcher.bsky.socialWe're excited about opportunities to track non-stationary variation in real-world systems, in extending beyond the many current analyses that assume stationarity.
Preprint again: arxiv.org/abs/2609.01651
@bendfulcher.bsky.socialOn sleep data, we show that non-stationary tracking methods can recapitulate a rich continuous fluctuation that broadly captures the manual categorical variation in sleep stage labels [developed in collaboration with Dr Tamaki (RIKEN)]
@bendfulcher.bsky.socialMany 'complex sounding' problems involving non-stationary chaotic attractors, are trivially solved by tracking low-order moments. And on the difficult ones, other state-of-the-art methods fail
@bendfulcher.bsky.socialThe key innovation lies in the combination of: (i) a comprehensive time-series feature set (to be sensitive to a wide range of sources of non-stationarity); but (ii) using SFA for dimension reduction, to extract only those features that vary slowly. [cf. PCA would not work]
@bendfulcher.bsky.socialNew preprint from Kieran:
We introduce a new data-driven method, feature-based slow feature analysis (f-SFA), to track non-stationarity from time-series data (with no generative model).
It beats benchmarks on chaotic systems and tracks sleep depth continuously from EEG.
arxiv.org/abs/2609.01651
@quantamagazine.orgReposted by @bendfulcher.bsky.socialMathematicians at OpenAI announced that a group of 10,000 autonomous AI agents under their direction proved that the Navier-Stokes equations can “blow up,” thus solving one of the six remaining Millennium Prize Problems. The result is not without controversy. www.quantamagazine.org/ai-has-solve...
@bendfulcher.bsky.social👏👏👏
@bendfulcher.bsky.social👏👏👏
@bendfulcher.bsky.socialNice good luck! Interested to hear how it goes for you
@bendfulcher.bsky.socialMy best student experiences came from take-home problems I could think about for weeks. Hard to incentivize this outcome for today's students with AI to generate full solutions.
I gave a talk on a model that aims to 'save' take-home assignments:
🎥 www.youtube.com/watch?v=slg0...
@bendfulcher.bsky.socialAn opportunity to *join our group* as a postdoc researcher analyzing complex physical systems, and the time-varying data they produce (with applications to sleep)!
Join us @sydneyphysics.bsky.social on a great campus in a beautiful city!
Please share 🙂
