iiinova — where your digital self keeps evolving

Reposted by daniel saunders

Alexandre Andorra @alex-andorra.bsky.social · 4h

Such a pleasure to host my friends Bill Engels and Jesse Grabowski, to talk #GaussianProcesses and #AIworkflows no less! Enjoy, my dear Bayesians 🖖

022

Reposted by daniel saunders

Nathaniel Haines @natehaines.bsky.social · 03/06/2026

here is the full sequence 🤓

0203

daniel saunders @danielsaunders.bsky.social · 27/02/2026

at my job I’m expectedutilitymaxxing. I chatted with @alex-andorra.bsky.social on how to make it scalable and fast in an industrial context on top of the probabilistic programming language you already love, PyMC.

011

Reposted by daniel saunders

Michael "Shapes Dude" Betancourt @betanalpha.bsky.social · 03/12/2025

In particular the latter step is often rich with modeling opportunities even if the theory of the latent phenomena is lacking. I often meet people who claim there is not enough domain expertise for a bespoke model but then go on for ten minutes about the intricacies of their measurement process.

111

daniel saunders @danielsaunders.bsky.social · 28/11/2025

Some people are out here practicing the One and True Scientific Method (bayes on structural economic models) and they are kind of enough to write books about how to do it.

100

Reposted by daniel saunders

Jessica Hullman @jessicahullman.bsky.social · 17/10/2025

It’s become fashionable in some circles to reject decision theory (and other basic statistical ideas) for vaguely political reasons. There are valid critiques—but also some worth being wary of. Some thoughts: statmodeling.stat.columbia.edu/2025/10/17/s...

statmodeling.stat.columbia.edu

Separating the whack from the chaff in critiques of decision theory | Statistical Modeling, Causal Inference, and Social Science

0154

Reposted by daniel saunders

British Journal for the Philosophy of Science @thebjps.bsky.social · 06/10/2025

Are you a phd student? Are you in the vicinity of one and want to help them with their bad life choices? Then this is for you! #philsci

1117

daniel saunders @danielsaunders.bsky.social · 13/08/2025

This is the classic shorter piece mechanism.ucsd.edu/bill/teachin... and he has a book length treatment of similar ideas in Representing and Intervening.

mechanism.ucsd.edu

020

daniel saunders @danielsaunders.bsky.social · 13/08/2025

Ian Hacking and Hasok Chang are big proponents of thinking about philosophy of science in terms of actions and skills. I love this paper by Chang: d1wqtxts1xzle7.cloudfront.net/52270296/201...

d1wqtxts1xzle7.cloudfront.net

140

daniel saunders @danielsaunders.bsky.social · 01/07/2025

Woo congrats to Nathan and your team! Excited to read this :)

010

daniel saunders @danielsaunders.bsky.social · 26/06/2025

Okay but what about when you are just trynna have a good time making stuff?

000

daniel saunders @danielsaunders.bsky.social · 04/06/2025

It’s an amazing package. The website also has a bunch of neat strategies to diagnose hmc samplers, ones I haven’t seen discussed elsewhere.

030

Reposted by daniel saunders

Julia M. Rohrer @dingdingpeng.the100.ci · 22/05/2025

“We didn’t ever hide that that’s what it was. People were mad because we were calling them effects,” she says. “Then they say to us, but they’re just associations with 20 covariates. But the point is we said that from the beginning. They’re associations with 20 covariates.”

1101

Reposted by daniel saunders

lastpositivist.bsky.social @lastpositivist.bsky.social · 22/05/2025

Friend telling me about how few statisticians are actually Bayesians. It's a real shame how far we have fallen from God's light.

5544

daniel saunders @danielsaunders.bsky.social · 13/05/2025

Insanely bad. I manually turn off the tab function whenever I’m about to be notebooking for a while.

000

Reposted by daniel saunders

Aki Vehtari @avehtari.bsky.social · 05/05/2025

Mitzi Morris's new case study mc-stan.org/learn-stan/c... illustrates with hierarchical and spatial models the better efficiency of the new sum_to_zero_vector constrained parameter introduced in Stan 2.36 (2024-12). Mitzi used CmdStanPy, but the Stan code is the same with all interfaces

mc-stan.org

The Sum-to-Zero Constraint in Stan

0194

daniel saunders @danielsaunders.bsky.social · 13/02/2025

It’s out 😮 I just checked last week to be teased with a mere abstract.

000

daniel saunders @danielsaunders.bsky.social · 15/01/2025

A brief history of tech disruptions: 2010: We are going to disrupt that horrible corporation, Blockbuster 2012: We are going to disrupt those awful cab monopolies 2014: We are going to disrupt the record labels 2022: We are going to disrupt reading and writing

000

daniel saunders @danielsaunders.bsky.social · 30/12/2024

Feels extremely true, watching how business leadership reacts to the long search process involved in finding a good model. The bayesian workflow literature largely assumes academic contexts where research papers are expected to be in development for a year or more.

000

Reposted by daniel saunders

PyMC Labs @pymc-labs.bsky.social · 24/12/2024

🎄✨ 𝐌𝐞𝐫𝐫𝐲 𝐂𝐡𝐫𝐢𝐬𝐭𝐦𝐚𝐬 𝐚𝐧𝐝 𝐇𝐚𝐩𝐩𝐲 𝐇𝐨𝐥𝐢𝐝𝐚𝐲𝐬 𝐟𝐫𝐨𝐦 𝐏𝐲𝐌𝐂 𝐋𝐚𝐛𝐬! 🎁 This holiday season, we want to thank everyone in our community for your support and enthusiasm. We’re grateful to see so many of you using PyMC-Marketing and CausalPy #MerryChristmas #HappyNewYear #PyMCMarketing #CausalPy #Gratitude

044

daniel saunders @danielsaunders.bsky.social · 25/11/2024

Sfu! bsky.app/profile/hkan... Also York! Canada has really strong terminal MAs.

010

daniel saunders @danielsaunders.bsky.social · 21/10/2024

briefly looking up empirical papers, giving up and resorting to nature documentaries is just ... extremely philosophy.

000

Reposted by daniel saunders

Paul Smaldino @psmaldino.bsky.social · 02/10/2024

Evolution of Similarity-Biased Social Learning (by me + Alejandro Pérez Velilla). Now in press at Evolutionary Human Sciences. Formalizes the long-standing idea that, for better or worse, it may be adaptive to ignore or down-weight information from outgroup sources. osf.io/preprints/so...

2289

daniel saunders @danielsaunders.bsky.social · 06/09/2024

It seems Ludwig Boltzmann had a bit of a drinking problem.

010

Reposted by daniel saunders

Nathaniel Forde @nathanielforde.bsky.social · 13/06/2024

New Deep Dive on Splines and Hierarchical Splines for modelling Insurance Loss curves with Bambi/PyMC. The focus is on the contrast between interpolation, extrapolation and how including extra hierarchical structure aids generalisation. nathanielf.github.io/posts/post-w...

171

Reposted by daniel saunders

Julia M. Rohrer @dingdingpeng.the100.ci · 14/04/2024

This is an excellent (very short!) discussion of how to decide which methods to use. (How can there be so many snappy and highly relevant pieces by Gelman et al. that I haven’t read?!)

2122

daniel saunders @danielsaunders.bsky.social · 28/03/2024

I picked up the dialectical biologist at a used bookstore a couple weeks ago. It slaps. Includes a whole chapter about pranks they played on EO Wilson along with this strategy for riches and professional success.

010

daniel saunders @danielsaunders.bsky.social · 20/03/2024

You might checkout what nathanielf.github.io does. He was a modal logic guy who now does causal inference in industry. Or www.pymc-labs.com/team/benjami... , used to be a cognitive science prof. Both of them would be pretty curious about your work I reckon.

010

daniel saunders @danielsaunders.bsky.social · 19/03/2024

That truly sucks! I’m so sorry the job market treated you this way. Have you explored applied causal inference consulting for industry?

100

daniel saunders @danielsaunders.bsky.social · 12/03/2024

Extremely cool! I’m looking forward to digging into this paper more!

000

daniel saunders @danielsaunders.bsky.social · 11/03/2024

I spend the first 5 weeks teaching python and simulation-based inference. Lots of emphasis on playing with model assumptions. No textbook. I just have all the lectures written up as colab notebooks github.com/daniel-saund...

000

daniel saunders @danielsaunders.bsky.social · 11/03/2024

I've been teaching a version of this class for a couple years. The hard tradeoff is how much of the class should be spent on practical skills vs explaining why common practice in journals is bananas. Students like the class more when there are more dunks obviously!

110

daniel saunders @danielsaunders.bsky.social · 01/03/2024

With blackjax, nutpie, numpyro, it compiles for 60 seconds. Then the progress bar goes from 0 to 100% in the blink of an eye. I'm telling you, people enjoy horse races more than formula 1 for a reason.

000

daniel saunders @danielsaunders.bsky.social · 01/03/2024

I'm starting to worry that the advanced compiler-NUTS samplers that have come out in the last couple years have taken the thrill out of bayes. In pymc3, I'd stare at the progress bar with glean and horror. Sometimes, it would get abruptly faster for no reason. Sometimes, it would stop.

000

daniel saunders @danielsaunders.bsky.social · 22/01/2024

Dropping by a Bayesian ecology conference today to see what they are doing. Apparently, it is strapping an accelerometer onto wild sharks. Then fitting hidden markov models to classify the sharks into "resting", "methodical exploration", and "rapid exploration".

010

daniel saunders @danielsaunders.bsky.social · 22/01/2024

Oh my. That’s a really neat trick.

010

daniel saunders @danielsaunders.bsky.social · 21/01/2024

Is the complaint half right though? Not the epidemiologists don’t think dynamically bit. But DAGs just cannot represent feedback or self-loops so it does seem funny to represent complex systems with them.

110

daniel saunders @danielsaunders.bsky.social · 17/01/2024

For sure, nuance always matters and toy models are more suggestive than conclusive. I just think this line of reasoning shows how fruitful the group selection framework can be - we know a lot about when it fails (migration) and how to patch up the failures (holding your labs mates accountable)

010

daniel saunders @danielsaunders.bsky.social · 17/01/2024

Some of these group-selection for cooperation arguments are designed to address this sort of worry. The groups maintain cooperation by punishing defectors, like a research group telling someone their lazy method won’t fly on this team.

000

Reposted by daniel saunders

Michael "Shapes Dude" Betancourt @betanalpha.bsky.social · 15/01/2024

Curious about how probability distributions transform? Ever wondered from where the term “marginal probability” comes? Confused by those “Jacobian” things? Have I got some writing for you.

342

daniel saunders @danielsaunders.bsky.social · 18/12/2023

Nancy Cartwright is so impressive. She has a MasterClass now. Even more impressive its on a subject not in her AOS.

Photograph of the actor Nancy Cartwright, not the philosopher.

140

Reposted by daniel saunders

Diane Coyle @dianecoyle1859.bsky.social · 17/12/2023

When preferences change: www.enlightenmenteconomics.com/blog/index.p...

274

daniel saunders @danielsaunders.bsky.social · 13/12/2023

when u open the assignment to see those overleaf-leave-space-for-the-reader-to-draw-graphs-sized margins and the default-I-know-my-IQ-score-type font, u can just assign them an A and move onto the next one.

000