Lightnews — Scholar-powered news

Reposted by Bruno Rodrigues

Carlos Scheidegger @cscheid.net · 22h

This may just be the best CS paper I’ve read this year. Just read the abstract and first para of the intro! The rest of the intro is really wild too, but very very good:

dl.acm.org/doi/pdf/10.1...

A screenshot of an academic paper. It reads:

Abstract
A "
'quine" is a deterministic program that prints itself. In this essay, I will show you a "gauguine": a probabilistic program that infers itself. A gauguine is repeatedly asked to guess its own source code. Initially, its chances of guessing correctly are of course minuscule. But as the gauguine observes more and more of its own previous guesses, it detects patterns of behavior and gains information about its inner workings.
This information allows it to bootstrap self-knowledge, and ultimately discover its own source code. We will discuss how-and why-we might write a gauguine, and what we stand to learn by constructing one.
CCS Concepts: • Computing methodologies → Philo-sophical/theoretical foundations of artificial intelli-gence; Theory of mind.
Keywords: reflection, probabilistic programming
ACM Reference Format:
Kartik Chandra, Amanda Liu, Jonathan Ragan-Kelley, and Joshua B.
Tenenbaum. 2025. Gauguin, Descartes, Bayes: A Diurnal Golem's Brain. In Proceedings of the 2025 ACM SIGPLAN International Symposium on New Ideas, New Paradigms, and Reflections on Programming and Software (Onward! '25), October 12-18, 2025, Singapore, Singa-pore. ACM, New York, NY, USA, 9 pages. https://doi.org/10.1145/
3759429.3762631

1 A Way of Knowing

From time to time, we all have crises of identity-moments of radical and overwhelming uncertainty about our selves.
I' don't know whether the doubts that seize us can really be externalized in language, but if I were to try, I would express them as questions, questions like: Who am I? What am I?
What kind of person? What kind of mind?

9 67 230

Bruno Rodrigues @brodriguesco.bsky.social · 6d

I think the reason is that file system is the natural predator of the data scientist. At least with jupyterlab, data scientists are able to find where they've saved their files more easily.

Bruno Rodrigues @brodriguesco.bsky.social · 7d

data scientists think mixing data, business logic and results in a single excel file is disgusting while they don't have an issue when they mix data, business logic and results in a single .ipynb file

curious

2

Bruno Rodrigues @brodriguesco.bsky.social · 7d

data scientists think mixing data, business logic and results in a single excel file is disgusting while they don't have an issue when they mix data, business logic and results in a single .ipynb file

curious

1 1 12

Bruno Rodrigues @brodriguesco.bsky.social · 7d

let's —and I cannot stress this enough— FUCKING GOOOO

1 2

Reposted by Bruno Rodrigues

CRAN Package Updates Bot @cranberriesfeed.bsky.social · 7d

New CRAN package rixpress with initial version 0.10.0
#rstats
https://cran.r-project.org/package=rixpress

CRAN: Package rixpress

Streamlines the creation of reproducible analytical pipelines using 'default.nix' expressions generated via the 'rix' package for reproducibility. Define derivations in 'R', 'Python' or 'Julia', chain them into a composition of pure functions and build the resulting pipeline using 'Nix' as the underlying end-to-end build tool. Functions to plot the pipeline as a directed acyclic graph are included, as well as functions to load and inspect intermediary results for interactive analysis. User experience heavily inspired by the 'targets' package.

cran.r-project.org

1 2 4

Reposted by Bruno Rodrigues

econmaett @econmaett.github.io · 8d

Personally I‘ve struggled with @docker.com containers because I don’t normally use Linux.

If I‘m not mstaken, {rix} by @brodriguesco.bsky.social should take care of all dependencies and might be an alternative.

r-consortium.org/posts/reprod...

Reproducibility Without Containers: Bruno Rodrigues Introduces Nix and the {rix} R Package – R Consortium

How can we ensure long-term reproducibility across operating systems, platforms, and time? Bruno Rodrigues believes the answer lies in Nix—and a package he developed for the R ecosystem.

r-consortium.org

1 5

Bruno Rodrigues @brodriguesco.bsky.social · 8d

yep it was and it's going to be published soon!

1 4

Bruno Rodrigues @brodriguesco.bsky.social · 13d

cool I'll be talking tomorrow at the LIH as well! www.lih.lu/en/event/%f0...

🇬🇧 Best practices for doing science with data » Luxembourg Institute of Health

🇬🇧 Best practices for doing science with data Speakers Head of Statistics Department – Ministry of Research and Higher Education, Luxembourg Abstract Data scientists and researchers write a lot of code but often without the practices that make it robust and reproducible. That gap can lead to serious consequences, such as the infamous Reinhart-Rogoff Excel […]

www.lih.lu

1 2

Reposted by Bruno Rodrigues

Peter Tennant @pwgtennant.bsky.social · 13d

Very excited to be in Luxembourg - with the Aging, Cancer, & Disparaties Research Unit green mascot - to launch the RESCOM lecture series in causal inference!

Future speakers from an amazing lineup include @georgiatomova.bsky.social, @miguelhernan.org, Mats Stenrund, & @dingdingpeng.the100.ci!

5 8 30

Bruno Rodrigues @brodriguesco.bsky.social · 19d

Same for rixpress: every step runs in a completely fresh and hermetic environment thanks to the power of Nix

1

Bruno Rodrigues @brodriguesco.bsky.social · 19d

interesting, users *should not* have to add miktex explicitely. Could you share the script that generated the default.nix

Bruno Rodrigues @brodriguesco.bsky.social · 19d

if you also have suggestions about UX or suggestions about how to make it easier to start as a user, please let me know!

1

Bruno Rodrigues @brodriguesco.bsky.social · 19d

Just tried it, and nope, help files work and show in the terminal! Can you share the dockerfile + default.nix? perhaps there’s something i’m missing.
Those warnings are unrelated.

1

Bruno Rodrigues @brodriguesco.bsky.social · 19d

weird, using a Nix shell shouldn't impact the help files! I'll try to reproduce.

1

Bruno Rodrigues @brodriguesco.bsky.social · 19d

ooooooor: docs.ropensci.org/rix/

3

Bruno Rodrigues @brodriguesco.bsky.social · 19d

ryxpress, a #Python package for building reproducible pipelines using #Nix as the underlying build tool: b-rodrigues.github.io/ryxpress/

ryxpress

Documentation for ryxpress

b-rodrigues.github.io

Bruno Rodrigues @brodriguesco.bsky.social · 20d

Friends, I’ll be giving a lecture titled "Best practices for doing science with data" at the Luxembourg Institute of Health next week (online as well, don’t hesitate to register): www.lih.lu/en/event/%f0...

🇬🇧 Best practices for doing science with data » Luxembourg Institute of Health

🇬🇧 Best practices for doing science with data Speakers Head of Statistics Department – Ministry of Research and Higher Education, Luxembourg Abstract Data scientists and researchers write a lot of code but often without the practices that make it robust and reproducible. That gap can lead to serious consequences, such as the infamous Reinhart-Rogoff Excel […]

www.lih.lu