A new paper from the lab!
We use MEG and a "local/global" design in the language domain to ask whether the transitions between words in a sentence are encoded by a shallow transition-probability mechanism, in parallel to a tree-based syntactic mechanism.
www.sciencedirect.com/science/arti...
It is unfortunate that the two types of model are often conflated, and that we do not discuss RG until graduate courses. Are there any undergraduate texts where RG is foregrounded?
The word “model” hides two different ideas.
Some models simplify reality to isolate a mechanism.
Others predict universal properties shared by many systems.
The renormalization group revealed the second possibility, and explains why simple models can describe Nature at all.
In the Casimir effect, the force depends only on an energy difference. Zee's book has a simple explanation (naive derivation assumes perfect infinite conductors) and Jaffe has all the details. But for the cosmological constant, there be dragons!
Published: Noise equals control.
I detail an explicit mapping between stochastic systems and control theory, showing how goal-directed behaviour arises from purely causal stochastic systems. Illustrated with stochastic resonance & Brownian ratchets
journals.aps.org/pre/abstract...
Depends on what synthesis you consider crucial but we found this paper useful:
pubmed.ncbi.nlm.nih.gov/33672143/
The arXiv preprint is updated to a full-length article (10 pages + 7 pages SI).
I will be around at StatPhys and Molveno if you want to chat!
( I am giving a talk about random languages in a ML session )
The talk is based on my recent preprint
arxiv.org/abs/2503.15670
written in an abbreviated manner for theoretical physicists. I am expanding the preprint to full article length and will update the arXiv soon.
In the talk I explain it all in simple language, and reinterpret stochastic resonance & Brownian ratchets with the noise-control mapping.
It also solves a conceptual problem in the field theory of stochastic systems, which is the physical interpretation of the `momentum' field that is introduced as a Lagrange multiplier, or via other abstract methods. The field is a control strategy. Agency is part of Nature
This solves the conceptual `agency-body' problem of dualism, i.e. reconciling a mechanistic description of biology with an agential one
My talk from
@bioctrl.bsky.social
about the physical origin of agency is uploaded here:
youtube.com/watch?v=YY21...
TLDR: noise equals control. Nature is constantly sampling control trajectories via noise
@philipcball.bsky.social potentially of interest! This result is very aligned with your perspective, I believe
I'm very excited about this virtual talk, tomorrow at 11am ET, and open to everyone! I will explain precisely how and why agency is part of Nature, and illustrate the idea with some examples from biology and beyond. No complex mathematics.
