Thank you, Ben! feel free to reach out if you have any Qs - and enjoy Cambridge!

Thank you, Henry! The short answer is no - although the reservoir is randomly initialized, the interpreted features are remarkably stable. High-dimensional expansion ensures key info is captured regardless of specific weights

Thank you Eric!

GitHub - rl671/heterorc: HeteroRC: Decoding latent information from dynamic neural responses with interpretable heterogeneous reservoir computing HeteroRC: Decoding latent information from dynamic neural responses with interpretable heterogeneous reservoir computing - rl671/heterorc

Code (python) is available at github: github.com/rl671/heterorc
Should be readily applicable to your own EEG/MEG decoding pipelines!
Great collaboration with @alexwoolgar.bsky.social @rhens.bsky.social , Sichao, Yanan, and John, during my time at @mrccbu.bsky.social & @theneuro.bsky.social ! (6/6)

In an attentional priority task (Duncan et al., 2023), HeteroRC uncovers statistically learned spatial priority information that remains ‘hidden’ from conventional methods, successfully decoding latent states previously thought to be ‘silent’. (5/6)

In a motor imagery dataset, it substantially improves decoding acc. and cross-temporal generalisation, revealing dynamic representational transformations. Also, its interpretation module opens the black box - linking decoding results to temporal, spectral, and spatial neural dynamics.(4/6)

In simulations, we found traditional linear decoders (SVM/LDA) are mainly sensitive to phase-locked evoked potentials. In contrast, HeteroRC robustly decodes information from evoked response, induced oscillatory power, inter-site phase synchronization (ISPC), and aperiodic spectral modulations.(3/6)

HeteroRC (Heterogeneous Reservoir Computing) projects neural signals into a high-D recurrent state space with different time constants, enabling nonlinear&multiscale decoding from raw signals. Reservoir weights are fixed - only the linear readout is trained, making it efficient & lightweighted (2/6)

Using time-resolved EEG/MEG decoding?🧠 Here’s a new approach!
No feature engineering (decode from raw signals), but capturing info that standard decoding often misses (oscillatory/aperiodic activity, connectivity).
Lightweight, INTERPRETABLE, and easy to use. (1/6)
www.biorxiv.org/content/10.6...

***Please forward to interested colleagues***

The call for symposia (deadline 15 April) and abstracts (deadline 1 May) for Biomag 2026 in Beijing (23-25 Aug) is now open:

biomag2026.scimeeting.cn

😐

The neuromodulatory fragility hypothesis of Alzheimer's disease pathogenesis Sporadic Alzheimer's disease (AD) is associated with numerous risk factors, yet its precise cause remains unclear. Here, we describe a novel framework for AD pathogenesis, whereby diverse risk factor...

Neuromodulatory neurons are extremely susceptible to stress. I argue that their fragility is an inherent weak point, the breaking of which directly leads to Alzheimer’s Disease.

I discuss this idea and its implications in a recently published perspective piece:

dx.doi.org/10.1002/alz....

Soon hiring a lab manager! Looking for someone who is really interested in language neuroscience, who is organised, motivated, a great communicator, and who works well in a research team. Express interest by submitting this form: tinyurl.com/glysn-labman...

Reposts appreciated!

Great work by Roni Tibon (not on BlueSky) - surprising that negligible difference in fMRI correlates of semantic vs episodic retrieval?

Aperiodic 1/f noise drives ripple activity in humans - Nature Communications How aperiodic 1/f noise drives ripple activity in human brain and impacts on ripple detections is not fully understood. Here authors show that ripple detections should be driven by the 1/f noise, whic...

Ripple oscillations are central for memory and sleep.

But ripple detection in humans remains challenging. Here we introduce a simulation approach in @natcomms.nature.com as common ripple detectors mainly pick up 1/f noise and not genuine oscillations

👇
www.nature.com/articles/s41...

#neuroskyence

Wow, finally!! 🎉🎉

WARN-D machine learning competition is live » Eiko Fried If you share one single thing of our team in 2026—on social media or per email with your colleagues—please let it be this machine learning competition. It was half a decade of work to get here, especi...

After 5 years of data collection, our WARN-D machine learning competition to forecast depression onset is now LIVE! We hope many of you will participate—we have incredibly rich data.

If you share a single thing of my lab this year, please make it this competition.

eiko-fried.com/warn-d-machi...

various computational neuroscience / MEEG / LFP short courses and summer schools

📆 updated for 2026!

list of summer schools & short courses in the realm of (computational) neuroscience or data analysis of EEG / MEG / LFP: 🔗 docs.google.com/spreadsheets...

New paper in Imaging Neuroscience by Ajay D. Halai, Marta M. Correia, et al:

Comparing the effect of multi-gradient echo and multi-band fMRI during a semantic task

doi.org/10.1162/IMAG...

BOLD signal changes can oppose oxygen metabolism across the human cortex, Nature Neuroscience

fMRI signals “up,” but neural metabolism might be going “down.”

In our @natneuro.nature.com paper, we demonstrate that about 40% of voxels with robust BOLD responses exhibit opposite oxygen metabolism, revealing two distinct hemodynamic modes.

rdcu.be/eUPO8
funds @erc.europa.eu
#neuroskyence 🧵:

How wonderful! Huge congrats, Alex!

How distributed is the brain-wide network that is recruited for cognition? - Nature Reviews Neuroscience Both localized and distributed views on the functional organization of the brain have been put forward. In this Perspective, Rosen and Freedman examine the degree to which these two views account for ...

How distributed is the brain-wide network that is recruited for cognition? A Perspective by Matthew C. Rosen & David J. Freedman

#neuroscience #neuroskyence

www.nature.com/articles/s41...

New paper in Imaging Neuroscience by Runhao Lu, Elizabeth Pollitt & Alexandra Woolgar:

Distinct and complementary mechanisms of oscillatory and aperiodic alpha activity in visuospatial attention

doi.org/10.1162/IMAG...

Please repost: another kind of Black-Friday deal!

most creative & interesting analysis!!

Special issue of Neuropsychologia celebrating the career of John Duncan.
www.sciencedirect.com/special-issu...
#neuroscience

We just published a paper in PNAS, showing microstructural changes to the hippocampus in aging and presymptomatic Alzheimer's disease - in humans, in vivo.

We continue to show the value of structural MRI beyond simply measuring large-scale atrophy!

Read here: ow.ly/1NvV50XkXnX

Why is cognitive effort experienced as costly? A widespread observation is that people avoid mentally effortful courses of action, and much recent work examining cognitive effort has explained subjective effort evaluation – and, consequently, pref...

New pontification piece with @awestbrook.bsky.social and Jean Daunizeau, just out in TICS:
Why is cognitive effort experienced as costly?
(or why does it hurt to think)

never written a review paper before in my life, that was a new and unusual experience

🚨New Results! Really excited to share this work -
Feedback doesn’t just signal right or wrong, it reorganizes population geometry of learning targets in vlPFC, moving representations toward future retrieval states. Notably, oscillations (TG-PAC & beta) gate when this transformation occurs.
🧵👇

BIOMAG 2025，在线会议，会议直播平台，网络会议直播，美迪康会务通，学术会议管理系统，会议活动管理，注册签到管理，会场幻灯片传输转播，PPT传输，会议网络直播，电子壁报管理系统，医学会，会员管理系统

Exciting news from the chairs of Biomag 2026, Prof. Jiahong Gao and Prof. Huan Luo — the conference website is now live: biomag2026.scimeeting.cn The meetings take place in Beijing, 23–25 August 2026. Save the date and start thinking about ideas for posters and symposia! Please share with colleagues