Ben Hardisty, PhD

Data mining & text mining on a scale you can't begin to fathom at Unknown. Lover of ☕️ 🥐🥓🐶 My normal is your never. #useR #Rstats #maths #T1D

[Facundo Miranda @Facundo59204769]

Hoy querria el handbook de oxford de filosofia de la fisica donde se encuentran articulos de varios temas fundamentales de filosofia de la fisica como la ontologia del espacio-tiempo hasta algunos muy poco explorados dentro del area como la filosofia de la hidrodinamica.

[Dr. Carmen Falcone @CarmenFalcone6]

The first work of the #FalconeLab is finally out!!! 🤩

A story of #astrocytes from evolution to disease. Thanks everyone who made this possible!

Varicose projection astrocytes: Conserved reactive cells in brain pathology | @ScienceAdvances https://t.co/BzNbnveFlO

@TowsonU

[SIAM Activity Group on Dynamical Systems @DynamicsSIAM]

Memorial Tribute: "A Legacy in Number Theory: A Tribute to Takashi Ono" (by Brian Conrad, Ken Ono, and Jacob Tsimerman; in Notices of the @amermathsoc): https://t.co/21q3IphoHe

[SIAM Activity Group on Dynamical Systems @DynamicsSIAM]

"Lecture notes on random matrix theory: the results, the applications, and the analytical tools" (by Joseph W. Baron): https://t.co/iUjn2qSwWZ

[Prachee Avasthi @PracheeAC]

A frustration of the AI era is trying to automate things that take more time than you’d like and are super easy for you (due to your experience and expertise) but less obvious for novices. Getting AI help on hard-for-you things feel like a jet pack/level up. Trying to automate the easy stuff is like smashing your face against the wall repeatedly.

And that’s when you realize your hard shit is someone else’s easy shit and I’m pretty sure collective frustration in the universe is going to increase until we hit ASI

[Physical Review E @PhysRevE]

#DYK: the spread of new words behaves like complex physics?

This paper introduces a piecewise power-law model that captures online language adoption, describing growth time-series, including sub-exponential patterns.

Read more: https://t.co/D9uwKB5asf

[Pixel Symphony @Pixel0Symphony]

Erika Giovanna Klien, Eggbeater. 1935.

Watercolor over graphite on tan wove paper.

[SIAM Activity Group on Dynamical Systems @DynamicsSIAM]

"2026 July Prize Spotlight" (in @TheSIAMNews): https://t.co/gag5Ns8zyb

[Prachee Avasthi @PracheeAC]

I’m extrapolating of course that he doesn’t mean that we should ban AI content when it’s better than what we allow from humans or beyond what we check for humans — and that the heuristics/merhodology we relied on don’t hold

[Prachee Avasthi @PracheeAC]

The AI-hesitant may be pushing to reverse a central value of arXiv to reflexively increase gatekeeping, but its founder Paul Ginsparg seems to recognize the issue more clearly:

“It’s now difficult to prepare for the world three months from now if the median LLM-produced computer science paper is better than that produced by the median grad student.”

https://t.co/ABZA3frZLZ

[Prachee Avasthi @PracheeAC]

This is a golden age of science. Some are doing society a service by automating every idiotic conception and flawed logic of human-constrained science/its historical contingencies, thereby making it more clear where the real opportunities lie to reshape research for the modern era.

[Quanta Magazine @QuantaMagazine]

The state of oceans and ice sheets — and, in turn, the climate — in 2100 will depend heavily on where they are right now. https://t.co/AB08Alw4wa

[PRX Life @PRX_Life]

While most models of the E.coli chemosensory lattice rely on allostery, this new paper presents an alternative approach that hinges primarily on zero-order ultrasensitivity and may better align with some qualitative experimental results.

Read more: https://t.co/1WxdUk3Icd

[Jeffrey West @mathoncbro]

@fpradelli94 describes his AI workflow to assess the relevance of each keyword-matched paper to our subject area of interest (mathematical modeling + cancer).

He attempts LLM classification with, and without the context of ~35 recent reviews in the field, & checks the agreement with human validation.

[Jeffrey West @mathoncbro]

One effective way to use Artificial Intelligence is to benchmark it within your own field.

That's exactly what we did, for mathematics in oncology:
https://t.co/GCpg4AKoBD

[Jeffrey West @mathoncbro]

@fpradelli94 This gives us a very accurate depiction of the field over time. We also found the earliest known examples of mathematical oncology -- two papers from 1886!

[PRX Life @PRX_Life]

A new framework that gauges the impact of ribosome-targeting antibiotics on mRNA translation provides testable hypotheses of mRNA pausing dynamics, and could inform the broader modeling of other biological transport processes.

Learn more: https://t.co/D2I7BgX5dt

[PRX Life @PRX_Life]

By introducing concepts analogous to overfitting and underfitting, this study shows how overfitness and underfitness impact selected population complexity — offering an explanation for how organism complexity evolves to optimally match their environment.

https://t.co/zFG53Pg5mW

[PRX Life @PRX_Life]

This study finds that Golgi matrix protein 130 is highly flexible and undergoes conformational transitions from extended chains to collapsed coils, providing insights on how structured proteins can support liquid-like, dynamic self-assembly.

🔗 https://t.co/6b8x0x5VzG

[PRX Life @PRX_Life]

When Physarum polycephalum is trapped inside polygons using harmful blue light, the #SlimeMold explores nearly the entire shape boundary until it identifies the most efficient contraction mode for its escape, usually via the polygon’s longest axis.

🔗 https://t.co/NZiqNebUju