Lightnews — Scholar-powered news

Manuel Baltieri

@manuelbaltieri.bsky.social

810 followers 300 following 130 posts

Chief Researcher at Araya, Tokyo. #ALife, #AI, embodied and enactive #cognition. Information, control and applied category theory for cognitive science. https://manuelbaltieri.com/

manuelbaltieri.com

Posts Media Videos Starter Packs

Pinned

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

After a multi-year collaboration with Martin Biehl, @mattecapu.bsky.social and @nathanielvirgo.bsky.social, I’m stoked to share the first of (hopefully) many outputs:
“A Bayesian Interpretation of the Internal Model Principle”
arxiv.org/abs/2503.00511.

1/

1 6 17

Reposted by Manuel Baltieri

Mariel Goddu @marielgoddu.bsky.social · Jul 11

My comment on Fillipo Torresan & @manuelbaltieri.bsky.social's "Disentangled representations for causal cognition" in Physics of Life Reviews:
www.sciencedirect.com/science/arti...

I argue that there is little meaningful analogy between learning from "pixels" vs "experience," but I praise

Disanalogies between causal learning in animals vs. machines: Comment on “disentangled representations for causal cognition” by F. Torresan & M. Baltieri

None.

www.sciencedirect.com

1 3 6

Manuel Baltieri @manuelbaltieri.bsky.social · May 2

Shocking

Manlio De Domenico @manlius.bsky.social · May 2

So, your favorite/fancy/rich AI provider most likely cheats most of the time to score high in LLMs leaderboards.

Shocking, but totally expected, isn't it?

arxiv.org/abs/2504.20879

Where are the AGI Bros? 👌

#AI #LLM

Reposted by Manuel Baltieri

theHigherGeometer @highergeometer.mathstodon.xyz.ap.brid.gy · May 2

Igor, you legend. Don't stop being you.

There are ten more of these unreadable hypercube diagrams on the following pages....

Souce: https://arxiv.org/abs/2505.00682

Two hypercube-shaped category-theoretic diagrams, each covered with an unreadable mess of labels.

1 4 1

Reposted by Manuel Baltieri

Anna Abalkina @abalkina.bsky.social · Apr 25

My experience applying for retractions at Elsevier.
I've looked at paper mills since 2019 and drafted a preprint on a paper mill from an international publisher in 2021. I started contacting journals or research integrity teams to raise concerns about papers. Publishers react differently. 1/n

2 15 30

Reposted by Manuel Baltieri

Johannes Kleiner @johanneskleiner.bsky.social · Apr 23

Great talk by @manuelbaltieri.bsky.social!

1 2

Reposted by Manuel Baltieri

Gordon @gordon.bsky.social · Apr 10

"We discuss the problem of running today’s software decades, centuries, or even millennia into the future" tinlizzie.org/VPRIPapers/t...

tinlizzie.org

2 5

Reposted by Manuel Baltieri

Fernando Rosas @frosas.bsky.social · Apr 8

Preprint time:
“AI in a vat: Fundamental limits of efficient world modelling for agent sandboxing and interpretability”
arxiv.org/abs/2504.04608

Exploring the fundamental limits that shape the design space of world modelling for agent sandboxing and interpretability

AI in a vat: Fundamental limits of efficient world modelling for agent sandboxing and interpretability

Recent work proposes using world models to generate controlled virtual environments in which AI agents can be tested before deployment to ensure their reliability and safety. However, accurate world m...

arxiv.org

1 6 13

Reposted by Manuel Baltieri

José A. Alonso @jalonso.bsky.social · Apr 2

GOL in GOL in HOL: Verified circuits in Conway's game of life. ~ Magnus O. Myreen, Mario Carneiro. arxiv.org/abs/2504.00263 #ITP #HOL4

GOL in GOL in HOL: Verified Circuits in Conway's Game of Life

Conway's Game of Life (GOL) is a cellular automaton that has captured the interest of hobbyists and mathematicians alike for more than 50 years. The Game of Life is Turing complete, and people have be...

arxiv.org

1 4

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 26

Great to see a colleague speaking up, sad to think about the state of affairs.

Reposted by Manuel Baltieri

ᛕᎥᕼᗷᗴᖇᑎᗴ丅Ꭵᑕᔕ @kihbernetics.bsky.social · Mar 23

I don't want to delete anything. I simply agree with Barbieri's distinction and claim that for a successful syntactic relationship, there is no need for anticipation or computation.
On that level, the cell is a simple reliable #state machine (transducer) with no place for interpretation of meaning.

1 1 1

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 15

📌

Reposted by Manuel Baltieri

Complexity Cat 🐱 @amahury.bsky.social · Mar 14

One of the most controversial corollaries of relational biology is the impossibility of simulating life. But what if I tell you that this claim is simply the result of misinterpreting Robert Rosen's ideas?

#complexitycat 🐈‍⬛👇🧵1/3

amahury.github.io/posts/trilog...

Relational Biology I: Is it possible to simulate life?

The first part of this trilogy is devoted to discussing the difference between model and simulation, one of the cornerstones for understanding relational biology. How true is it that Robert Rosen deni...

amahury.github.io

2 5 10

Reposted by Manuel Baltieri

Konrad Kording @kordinglab.bsky.social · Mar 12

Re the Tononi paper: Both Tononi’s IIT (phi) and Friston’s FEP start from fundamental, axiomatic, and debatable assumptions. These assumptions are generally made without any humility. This logic allows them to make exceptionally broad claims. Which contributes to my unease about them.

9 17 92

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 10

Directed wiring diagrams for Mealy machines!

arxiv math.CT @arxiv-math-ct.bsky.social · Mar 10

Keri D'Angelo, Sophie Libkind
Dependent Directed Wiring Diagrams for Composing Instantaneous Systems
https://arxiv.org/abs/2503.05457

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 8

📌

Reposted by Manuel Baltieri

lastpositivist.bsky.social @lastpositivist.bsky.social · Mar 8

Only just learning about this now -- I guess for a while people have predicted that the AI doomer rationalist crowd would go violent, so its not surprising in some sense. Still though, odd times!

www.theguardian.com/global/ng-in...

They wanted to save us from a dark AI future. Then six people were killed

How a group of Silicon Valley math prodigies, AI researchers and internet burnouts descended into an alleged violent cult

www.theguardian.com

10 15 92

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

Secondly, we discuss how this form of Bayesian filtering is quite simplistic, 1) not making full use of Bayesian updates by ignoring observations from the environment/plant, and 2) assuming that beliefs of equicredible states of the environment are disjoint (they form a partition).

16/16

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

Importantly, this makes use of the fact that we have a Markov category, Rel^+, of possibilistic Markov kernels that can be used to specify beliefs as (sub)sets without assigning them probabilities, but that works very much like other “nice” Markov categories.

15/

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

We then show how this corresponds to a Bayesian filtering interpretation for a reasoner: how a controller modelling its environment can be understood as performing Bayesian filtering on its environment.

14/

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

Firstly, we show that the definition of model between two autonomous system can be “reversed” to build a “possibilistic” version of the internal model principle.

13/

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

After a reasonably self contained overview of string diagrams for Markov categories, and some definitions including Bayesian inference/filtering, their parametrised and conjugate prior versions, we dive into the main result, showing mainly two things.

12/

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

In the second part of the paper, we use results from a recent line of work (link.springer.com/chapter/10.1...) started by some of my collaborators on how to interpret a physical system as performing Bayesian inference, or filtering, using the language of Markov categories.

11/

Interpreting Dynamical Systems as Bayesian Reasoners

A central concept in active inference is that the internal states of a physical system parametrise probability measures over states of the external world. These can be seen as an agent’s beliefs...

link.springer.com

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

Our focus here is mostly technical and has to do almost entirely with control theory, but considering where the conversation started on the other platform, I hope that this will have an impact also in the cognitive and life sciences.

10/

1 1

Manuel Baltieri @manuelbaltieri.bsky.social · Mar 5

This is often taken to be 1) a better formalisation of Conant&Ashby’s good regulator “theorem”, and 2) the reason why talking about “internal models” is necessary in cognitive science, AI/ML/RL, biology and neuroscience.

/9

1 2