Tomo Hiroyasu @HTechLab

Can exercise change how we feel negative emotions? 🏃‍♀️ Wang and Zhou (2026) show with #fNIRS that habitual exercisers have lower negative emotional reactivity, suggesting more efficient early emotional processing. Read more here: 📖 doi.org/10.3390/brains…

keyboard_arrow_left Previous keyboard_arrow_right Next

🚨 New Preprint! 🧠 We gave an AI model one simple rule: rearrange your neurons so that nearby ones respond alike. We never told it what a face, a voice, or a sentence was. It grew brain-like maps for all three anyway. 🧵👇 🌐 Website: topo-omni.epfl.ch

@Falkerisen @memeticsisyphus In yellow is a meta-analytic task fMRI map of reasoning it misses primary cortex (outlined in black) but it decently distributed. Reasoning is very dissimilar from the IQ-brain association map (green). However, the IQ association map is very similar to the arousal pattern.

🚨 Breaking neuroscience/evolution news! New Nature paper provides further evidence for a central hypothesis (with a long lineage) that I defended in my 2004 book, The Birth of the Mind: the evolutionary process of duplication and divergence played a key role in brain evolution.

What matters most for childhood brain organization? We analyzed 649 variables. The answer: Socioeconomics (SES); with brain patterns pointing at sleep & stress as drivers. Even brain-IQ associations were better explained by SES confounding. In Science: science.org/doi/10.1126/sc…

In today's @TheLancet there are 3 papers on cardiometabolic disease: biology, epidemiology, prevention/treatment. The sobering and all to common story from womb to tomb conveyed in this graphic thelancet.com/journals/lance… thelancet.com/journals/lance… thelancet.com/journals/lance…

New paper in Imaging Neuroscience by Tiara Bounyarith, David Braun and Aaron Kucyi: Real-time fMRI-triggered experience sampling: A proof-of-concept study doi.org/10.1162/IMAG.a…

I'm the brain, context is everything!

Julien Dirani @DiraniJulien

a month ago

Check out our new preprint! arxiv.org/abs/2605.23111 Contextual Role Modulates Object Representational Geometry in the Human Brain

keyboard_arrow_left Previous keyboard_arrow_right Next

2 16 56 10K 48

Better training data, not bigger: an entropy-symmetry map for machine-learning interatomic potentials Machine learning potentials (MLPs) let us run atomic simulations near DFT accuracy but far faster, reshaping how we study materials. Their weak point is data. Crystal configuration space explodes combinatorially, and most sampling either leans on idealized symmetric structures or skips the metastable, disordered states that drive real behavior. Collecting more configurations does not fix this, and uneven coverage quietly poisons predictions. Zeyuan Li and coauthors attack the problem with a clever low-dimensional descriptor. They project any configuration onto two physically meaningful axes: the instantaneous pair entropy (radial, thermodynamic order) and the sixth-order Steinhardt parameter (angular, crystallographic order). Unlike PCA or t-SNE, this "entropy-symmetry landscape" is interpretable and information-preserving, and distances on it reflect genuine structural similarity. On top of it they build two tools. PolymorphGen uses genetic-mutation sampling to generate polymorphs directly on the landscape, decoupling exploration from expensive energy evaluation, and even found a graphite-to-diamond pathway with a lower barrier than standard NEB. MLPKD then distills accuracy from DFT through a message-passing network down to a lightweight model, a million-fold speedup with little accuracy loss. The payoff is two scaling laws worth knowing. Under uniform coverage, accuracy saturates fast: as few as 77 configurations captured iridium's brittleness and phonon spectra. But biased coverage causes large extrapolation errors no matter the dataset size. Diversity and uniformity beat raw count. The distilled model then resolved twinning in iridium and dislocation slip in aluminum, where a 980,000-configuration active-learning baseline got iridium's elastic constants wrong. In materials, alloy, and energy R&D, the bottleneck in atomistic ML is rarely compute, it is how well your training set spans configuration space. A cheap, physics-grounded map like this flags where data is thin before you burn DFT cycles, shrinking simulation campaigns for new alloys, battery materials, or structural metals from months to days. Paper: Li et al., Nature Communications (2026) — CC BY-NC-ND 4.0 | doi.org/10.1038/s41467…

Our latest, out now in @NatMentHealth led by @IvaIlioska using normative modelling to understand the heterogeneity of atypical FC in autism across different resolution scales: nature.com/articles/s4422…

𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝘄𝗵𝗶𝘁𝗲 𝗺𝗮𝘁𝘁𝗲𝗿 𝘀𝘂𝗽𝗽𝗼𝗿𝘁 𝘀𝗶𝗴𝗻𝗮𝗹 𝗽𝗿𝗼𝗽𝗮𝗴𝗮𝘁𝗶𝗼𝗻 𝗶𝗻 𝘁𝗵𝗲 𝗯𝗿𝗮𝗶𝗻? How about traveling waves?? jneurosci.org/content/early/…

New paper in Imaging Neuroscience by Gabriel Riegner, Armin Schwartzman, et al: Estimating fMRI timescale maps doi.org/10.1162/IMAG.a…

Congrats to Longzhou @xlzhope on this exciting new work: "Long-range white-matter pathways enable efficient spontaneous neural activity propagation in the human brain", now out in J. Neurosci. @SfNJournals: jneurosci.org/content/early/…

SfN Journals @SfNJournals

a month ago

#JNeurosci: Findings from @xlzhope, @zaixucui et al. provide insight into how the brain supports efficient large-scale communication and link brain structure to fast neural dynamics. doi.org/10.1523/JNEURO…

0 17 37 9K 14

New article in @PNASNews: We all know that ChatGPT loves to delve, bolster, leverage, encompass, showcase, underscore, et cetera. I analyzed full text of 7.3 million journal articles published 2020-2025, hunting for 228 words that spiked after ChatGPT launched in late 2022.

[Aging] Akbaş et al.: "The brain-heart interconnection may alter through aging and cognitive impairment." doi.org/10.1016/j.clin…

keyboard_arrow_left Previous keyboard_arrow_right Next

Minogue & Zouridakis et al. studied the brains of people with dementia and FTLD-PSP or FTLD-CBD. They found that PSP and CBD show distinct patterns of selective vulnerability, with tau pathology predominantly affecting glia in PSP and neurons in CBD. shorturl.at/WPMzh

Don't miss our work presented at #ICRA2026 in Vienna 🇦🇹 📅 June 4, 12:00-12:10, Hall A2 "Congestion Mitigation Path Planning for Large-Scale Multi-Agent Navigation in Dense Environments" (RA-L) T Kato, KO, Y Sasaki & N Yokomachi arxiv.org/abs/2508.05253

Over at @Neuro_Skeptic, there's details of a "fun" new paper that looks at the changing landscape of #neuroimaging research. Check it out here: bit.ly/2NmCtuA #neuroscience

The dynamic epigenetic landscape of the retina and the accompanying cloud based data resource @StJude pecan.stjude.org/#/proteinpaint…

New paper in Imaging Neuroscience by SungJun Cho, Mark W. Woolrich, et al: Modelling discrete states and long-term dynamics in functional brain networks doi.org/10.1162/IMAG.a…