Sign in

Giulio Ruffini

@ruffini.bsky.social
409 followers 238 following 105 posts

Physicist working on computational neuroscience, brain stimulation and foundational aspects of (meta)physics, swimming and music during spare CPU cycles. Neuroelectrics.com, Starlab.es, BCOM.one

PostsRepliesMedia
Giulio Ruffini @ruffini.bsky.social · 11/09/2026
My take on the Hugging Face hack! www.giulioruffini.com/blogs/the-ag...
giulioruffini.com
010
Reposted by Giulio Ruffini
Ricard Solé @ricardsole.bsky.social · 11/09/2026
La IA como un virus cognitivo: comentario en @lavanguardia.com acerca de nuestra investigación reciente @ruffini.bsky.social @francesca-castaldo.bsky.social @sfelena.bsky.social @brigan.bsky.social www.lavanguardia.com/vida/2026091...
081
Reposted by Giulio Ruffini
Neuroelectrics @neuroelectrics.bsky.social · 10/09/2026
Can adapting brain circuitry help prevent suicide? On #WorldSuicidePreventionDay, every advancement in understanding and treating depression is a step forward in the fight against suicide. Learn more: youtube.com/watch?v=pqBc... #MentalHealth #Neuromodulation
youtube.com
Suicide Prevention by Adapting Brain Circuitry
YouTube video by Neuroelectrics
021
Giulio Ruffini @ruffini.bsky.social · 10/09/2026
Preprint update 👉 doi.org/10.5281/zeno... Now with more fun theorems and proofs (LEAN checked). @bcom-foundation.bsky.social @francesca-castaldo.bsky.social
doi.org
From "More Is Different" to Algorithmic Emergence: Regularity, Compression, and the Limits of Discovery
Scientific discovery seeks regularities that support explanation and prediction. Compression makes their reuse explicit: shared structure is described once, and parameters specify each case. A model c...
000
Giulio Ruffini @ruffini.bsky.social · 10/09/2026
Algorithmic emergence: the event where an observer/agent discovers a useful representation and reusable model that reveal regularity it could not use before. Compression may not exist—and even when it does, every reliable method that always halts has blind spots.
110
Reposted by Giulio Ruffini
Ricard Solé @ricardsole.bsky.social · 04/09/2026
Can AI spread like a virus and erode cognitive autonomy? Our new paper models how LLM adoption can cross tipping points, triggering runaway dependence and tech lock-in, suggesting routes to "cognitive immunization". arxiv.org/abs/2609.03344 @ruffini.bsky.social @francesca-castaldo.bsky.social
14816
Giulio Ruffini @ruffini.bsky.social · 14/07/2026
yes, it does, and it is a very interesting scenario!
100
Giulio Ruffini @ruffini.bsky.social · 13/07/2026
New preprint out! => The cortical column as a tuned receiver: a network mechanism for temporal-interference stimulation. We propose a mechanism for #TI leveraging oscillatory computation in the brain. How? Nonlinearity + resonant network amplification, like AM radio! 📻 doi.org/10.5281/zeno...
010
Giulio Ruffini @ruffini.bsky.social · 13/07/2026
then they get a particular kind of structured experience we call self-awareness. I believe Experience is fundamental, not to be explained by any more fundamental process.
010
Giulio Ruffini @ruffini.bsky.social · 13/07/2026
I think that what you refer to is a "self-model"? For me, that has to do with self-awareness. I don't like using the word "consciousness" because it is very overloaded. I like Experience and Structured Experience as the better-defined terms. If agent structure includes a self-model,
110
Giulio Ruffini @ruffini.bsky.social · 13/07/2026
In our new paper, "Pattern, Persist!", Francesca Castaldo and I converge on the "algorithmic agent" as the structure behind Persistence. We use this to reframe evolution, revisit the free-energy principle and thermodynamics, and address human-AI alignment. 👉 zenodo.org/records/2127...
zenodo.org
Pattern, Persist!
The word agent now names systems as different as a tool-using language model and a chemotactic cell, with no definition the fields that use it would share. We argue these uses converge on one substrat...
100
Giulio Ruffini @ruffini.bsky.social · 13/07/2026
What survives the Filter of Time? Will we? What do a proton, a flame, a chemotactic cell, and an LLM agent have in common? ⚛️🧬🐕🤖 (fast intro: giulioruffini.com/blog/) @bcom-foundation.bsky.social
211
Giulio Ruffini @ruffini.bsky.social · 06/07/2026
In spanish: giulioruffini.github.io/assets/blogs...
giulioruffini.github.io
Las cosas que permanecen — Blog de BCOM
000
Giulio Ruffini @ruffini.bsky.social · 06/07/2026
The things that stay. @bcom-foundation.bsky.social You cannot step into the same river twice—but the river remains. 🌊 From Aristotle to Alan Turing, how do we persist when our atoms are constantly replaced? #Philosophy #CognitiveScience #KT giulioruffini.github.io/assets/blogs...
141
Giulio Ruffini @ruffini.bsky.social · 28/03/2026
The transformer has no memory. Some thoughts on LLMs, memory, RAG, and how to use Claude. bcomone.atlassian.net/wiki/spaces/... @bcom_foundation @bcom-foundation.bsky.social @francesca-castaldo.bsky.social
141
Giulio Ruffini @ruffini.bsky.social · 16/03/2026
All algorithmic agents welcome here!
051
Giulio Ruffini @ruffini.bsky.social · 03/03/2026
5/5 Whether you are a single cell, a human brain, or an AI, to maintain homeostasis, you must run a generative model of your environment. You can't just react; you must predict! 🌍🤖 Check out the full Open Access paper here: 🔗 www.mdpi.com/1099-4300/28...
mdpi.com
100
Giulio Ruffini @ruffini.bsky.social · 03/03/2026
4/5 Why does this matter? It provides a rigorous, distribution-free mathematical backbone for neuroscience frameworks like the Free-Energy Principle and Active Inference.
mdpi.com
400
Giulio Ruffini @ruffini.bsky.social · 03/03/2026
3/5 The paper looks at regulation as data compression. It shows that if a regulator successfully keeps a system (embedded in the world) stable (reducing the algorithmic complexity of its output), it must share "mutual algorithmic information" with the world with high probability.
mdpi.com
110
Giulio Ruffini @ruffini.bsky.social · 03/03/2026
2/5 In a new paper, "The Algorithmic Regulator" (published in Entropy), I tackle this using Algorithmic Information Theory and Kolmogorov Complexity. It takes the classic Good Regulator Theorem and the Internal Model Principle and complements and extends them for non-linear, deterministic systems.
mdpi.com
130
Giulio Ruffini @ruffini.bsky.social · 03/03/2026
Do you need to understand the world to survive in it? A classic 1970 cybernetics theorem says yes: "Every good regulator of a system must be a model of that system." But proving this mathematically for complex, unpredictable real-world scenarios has always been notoriously difficult. 1/5
120
Giulio Ruffini @ruffini.bsky.social · 16/02/2026
Annotated version: oup.silverchair-cdn.com/oup/backfile...
oup.silverchair-cdn.com
000
Giulio Ruffini @ruffini.bsky.social · 16/02/2026
This is the paper where, after almost 20 years, I started taking my AIT obsessions a bit more seriously. I am happy I did... maybe you have some too... => Don't wast time! : ) academic.oup.com/nc/article/2... PS: the Supplementary Data part is more fun!
110
Giulio Ruffini @ruffini.bsky.social · 16/02/2026
Philip, agree on 1... but I think the road for studying consciousness is assuming primordial "experience" and instead focus on "structured experience" - and this connects directly with mathematics. See my talk in the "Platonic Space Symposium" thoughtforms.life/wp-content/u...
thoughtforms.life
000
Giulio Ruffini @ruffini.bsky.social · 16/02/2026
Thanks, Johannes!
000
Giulio Ruffini @ruffini.bsky.social · 15/02/2026
An Introduction to Galois Theory (with connections to AIT): zenodo.org/records/1845... ...This note aims to demystify Galois Theory by connecting its foundational definitions to a broader principle of computational and compositional tractability.
150
Giulio Ruffini @ruffini.bsky.social · 15/02/2026
From The Sorcerer's Apprentice to Crystal Nights: Security Implications from Moltbot/Moltbook to Greg Egan's Crystal Nights! zenodo.org/records/1844...
000
Reposted by Giulio Ruffini
Ricard Solé @ricardsole.bsky.social · 14/02/2026
Could life have begun with simpler molecules than we once thought? A new paper in @science.org by @edogia.bsky.social shows that a tiny RNA catalyst can self-replicate itself, suggesting that life may have been easier to emerge than expected. Getting closer. www.biorxiv.org/content/10.1...
14718
Giulio Ruffini @ruffini.bsky.social · 15/02/2026
Updated version with graphical materials now available here: zenodo.org/records/1864...
zenodo.org
The Algorithmic Regulator (final version and graphical resources)
The regulator theorem states that, under certain conditions, any optimal controller must embody a model of the system it regulates, grounding the idea that controllers embed, explicitly or implicitly,...
000
Giulio Ruffini @ruffini.bsky.social · 15/02/2026
Thanks for spreading the word, Ricard!
010
Reposted by Giulio Ruffini
Ricard Solé @ricardsole.bsky.social · 12/02/2026
What if we had a Rosetta Stone for brain oscillations—one framework to translate between models and scales? In this paper lead by F Castaldo and @ruffini.bsky.social they build a simple, systematic ladder of neural mass models showing how diverse formalisms connect arxiv.org/pdf/2512.10982
1195
Giulio Ruffini @ruffini.bsky.social · 31/12/2025
My annotated slides for my talk in the wonderful thoughtforms.life/symposium-on... organized by @drmichaellevin are here: giulioruffini.github.io/assets/slide...
010
Giulio Ruffini @ruffini.bsky.social · 19/12/2025
Thanks! That was straight from the paper! Uploaded it, clicked on slides…. That’s it!
010
Giulio Ruffini @ruffini.bsky.social · 19/12/2025
Paper: mdpi.com/1099-4300/27... Related: arxiv.org/abs/2510.10586 More on Algorithmic agents: giulioruffini.github.io
mdpi.com
000
Giulio Ruffini @ruffini.bsky.social · 19/12/2025
Impressed: #notebooklm created a nice presentation of the Algorithmic Agent and Symmetry paper! github.com/giulioruffin...
github.com
210
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
15/ ... plus, linearization of Wilson-Cowan; Connecting SL with Wilson-Cowan.
000
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
14/ Other goodies: discussion using L-operators (for synapses) and transfer functionals.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
Definition: A dataset is said to represent an oscillation when it can be most succinctly Lie-generated from a representation of U1 (plus noise). #ait #kolmogorov
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
12/ Bonus material: plenty of good stuff in the Appendix for aficionados, including links with Groups, Topology, and Algorithmic Information Theory (What is an oscillation)? @ERC_Research @neurotwin @Neuroelectrics
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
11/ If you use neural mass models (or teach them), I’d love feedback: what translation step is hardest in your workflow? PDF: arxiv.org/pdf/2512.10982 #computationalneuroscience #neuralmass #EEG #MEG
arxiv.org
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
10/ Practical cheat-sheet: • Phase locking/entrainment → phase models • Spectra/covariances → damped linear resonators • Limit cycles near Hopf → Stuart–Landau • Firing-rate E–I loops → Wilson–Cowan • PSP/synaptic kinetics → NMM1 • First-principles spiking link → NMM2
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
9/ NMM2 (next-generation masses): Exact mean-field reductions of QIF networks yield dynamic (r,v) equations—a *dynamic* transfer function replacing static sigmoids, linking spikes ↔ masses.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
8/ NMM1 (second-order synapses): PING-like motifs, Jansen–Rit, Wendling, and laminar neural masses become variants of one formalism— highlighting which parameters control resonance, PSPs, and phase shifts.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
7/ Wilson–Cowan then appears as the same backbone made explicit: an E–I push–pull loop + delayed, nonlinear transfer → Hopf/limit cycles, with clean recipes for forcing & coupling.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
6/ Interlude: synapses + transfer functionals. Synaptic filters create delays/phase lags; nonlinear transfer function(al)s map summed input → firing-rate output. We keep forcing/coupling consistent across model “dialects” via an operator viewpoint.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
5/ Each rung is treated in 3 regimes: (i) isolated node, (ii) forced node (external drive), (iii) coupled network. Because that’s how models meet data *and* interventions.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
4/ The undamped harmonic oscillator (phase) and the damped HO (phase and amplitude) are natural starting points. We climb a ladder: HO → damping/forcing/coupling → driven resonator → nonlinearity → Stuart–Landau (Hopf normal form) → connect to classic firing-rate & synapse-based models.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
3/ Neural mass models power EEG/MEG/fMRI generators, whole-brain simulations, and perturbation/stimulation studies—but the zoo of formalisms makes model choice & interpretation messy.
100
Giulio Ruffini @ruffini.bsky.social · 15/12/2025
2/ Our starting point is simple: Oscillations can be seen as a push–pull interaction between two effective degrees of freedom (think E↔I, or quadrature components). That core motif survives as we add biology.
110