iiinova — where your digital self keeps evolving

Reposted by Aki Vehtari

MC Stan @mc-stan.org · 17/09/2026

Release of CmdStan 2.40, including new Stan Math, core Stan, and Stanc, is now available! Blog post with new features blog.mc-stan.org/2026/09/16/r... Highlights - integrate_1d_double_exponential, a variadic version of old integrate_1d - integrate_1d_gauss_kronrod, a different quadrature rule ...

blog.mc-stan.org

Release of CmdStan 2.40

We are very happy to announce that the 2.40.0 release of CmdStan is now available on Github! As usual, the release of CmdStan is accompanied by new releases of Stan Math, core Stan, and Stanc3. Thi…

13016

Aki Vehtari @avehtari.bsky.social · 16/09/2026

"Recommendations for visual predictive checks in Bayesian workflow" published in Journal of Visualization and Interaction doi.org/10.54337/jov..., which is exciting also because this is our first Quarto html journal paper www.journalovi.org/2025-sailyno.... I wish more journals would adopt this!

doi.org

Recommendations for visual predictive checks in Bayesian workflow | Journal of Visualization and Interaction

0276

Reposted by Aki Vehtari

Aki Vehtari @avehtari.bsky.social · 08/09/2026

I'm looking for doctoral students, postdocs and research fellows to work with on Bayesian Workflow. Flexible starting time

02115

Reposted by Aki Vehtari

Martin Modrák @modrakm.bsky.social · 14/09/2026

I will be looking for a PhD student in Bayesian statistics (in collaboration with Paul Bürkner), located in Prague. Forward to students if you have somebody who could be interested. www.martinmodrak.cz/grammo-call/

martinmodrak.cz

Grammo - Let's Research Bayes!

03224

Aki Vehtari @avehtari.bsky.social · 09/09/2026

A FAQ is what to do in case of high Pareto-k's with PSIS-LOO or PSIS-LOGO. In case of hierarchical models we can integrate out group specific parameters before PSIS. We investigated reliability of some alternative integration methods with the eventual goal of automating this arxiv.org/abs/2609.05713

1233

Reposted by Aki Vehtari

Juha Karvanen @juhakarvanen.bsky.social · 09/09/2026

New paper in JMLR with Otto Tabell and Santtu Tikka: “Clustering and Pruning in Causal Data Fusion.” We derive conditions for reducing causal graphs while preserving conclusions about identifiability and non-identifiability. www.jmlr.org/papers/v27/2...

Three causal diagrams illustrating graph reduction. The first shows the original causal graph, the second shows a pruned graph after irrelevant variables are removed, and the third shows a clustered graph in which three variables are combined into one node labeled T.

0101

Aki Vehtari @avehtari.bsky.social · 08/09/2026

I'm looking for doctoral students, postdocs and research fellows to work with on Bayesian Workflow. Flexible starting time

02115

Reposted by Aki Vehtari

Jarno Vanhatalo @jarnovanhatalo.bsky.social · 08/09/2026

During the many long term ecological change projects in @rececochange.bsky.social, we have concretely learned how important long term data are. Here some thoughts on the topic 👇 Long-Term Community Data Reveal Ecological and Evolutionary Phenomena - www.annualreviews.org/content/jour...

annualreviews.org

Long-Term Community Data Reveal Ecological and Evolutionary Phenomena

Resolving the ecological and evolutionary processes affecting biodiversity requires long-term community data. By separating short-term variability from directional change and by revealing lags spannin...

0116

Aki Vehtari @avehtari.bsky.social · 07/09/2026

Xiang Ye's excellent Stan case studies for 1) new priors for circular distributions xiangyestats.github.io/directional-..., 2) and new models for circular targets xiangyestats.github.io/directional-... 3) and circular covariates xiangyestats.github.io/directional-... Links to papers included

xiangyestats.github.io

Penalized Complexity Priors for Circular Distributions

Construct and calibrate PC priors for von Mises concentration, with circular-uniform and point-mass base models.

0223

Aki Vehtari @avehtari.bsky.social · 04/09/2026

As non-native English speaker from Finland it was interesting to test this, but I'm not surprised by the result

380

Reposted by Aki Vehtari

Vincent Arel-Bundock @vincentab.bsky.social · 04/09/2026

Big news! 🎉 𝚖𝚊𝚛𝚐𝚒𝚗𝚊𝚕𝚎𝚏𝚏𝚎𝚌𝚝𝚜 1.0.0 for #Rstats is out. It’s a big number and it feels like a big step. I wrote a blog on the challenges of interpreting statistical models, SPEED, cool new features, the future, and a 5 year package development and writing odyssey. arelbundock.com/posts/margin...

1642097

Reposted by Aki Vehtari

Richard McElreath 🐈‍⬛ @rmcelreath.bsky.social · 02/09/2026

Last year, I publicly complained multiple times (see e.g. elevanth.org/blog/2025/07...) that the open access movement had been captured by publishers, and we are now worse off than before open access. Well people got mad at me for saying it. But I think things are still getting worse.

616466

Reposted by Aki Vehtari

Jarno Vanhatalo @jarnovanhatalo.bsky.social · 02/09/2026

We made a systematic comparison of contemporary remote sensing techniques (RS) and species distribution modelling (SDM) in predicting spatial distribution of coastal habitats in seven European marine areas. www.sciencedirect.com/science/arti... 1/3

sciencedirect.com

121

Reposted by Aki Vehtari

ELLIS Institute Finland @ellisinstitute.fi · 21/08/2026

Now #hiring: postdocs and PhD students in AI & machine learning research. Deadline Sept. 21, 2026. - Cutting-edge computational resources like @lumi-supercomputer.eu - Close collaboration with @ellis.eu network - Ambitious, high-impact projects across Finland www.ellisinstitute.fi/postdoc-and-...

21010

Reposted by Aki Vehtari

Visruth Srimath Kandali @visruth.com · 01/09/2026

`stanflow` v0.2.0 is out! `stanflow` is a metapackage a la `tidyverse` to provide a convenient way to install and manage #rstats Stan packages. The package is still in early stages; as such, if you want some features they can probably be added! Stan forum: discourse.mc-stan.org/t/stanflow-v...

discourse.mc-stan.org

`stanflow` v0.2.0 is out

It’s been a while, but stanflow v0.2.0 has just been released by @Visruth. stanflow is a metapackage a la tidyverse to provide a convenient way to install and manage R Stan packages. (See Soft relea...

0166

Reposted by Aki Vehtari

Juha Karvanen @juhakarvanen.bsky.social · 01/09/2026

When does“variance explained” actually mean causation? In our new preprint, Olli Saarela and I develop graph-based causal variance decompositions. The framework clarifies when components of an ordered variance decomposition have causal interpretations. arxiv.org/abs/2608.27140

arxiv.org

Graph-based causal variance decompositions: When "variance explained" means causation

Recursive application of the law of total variance decomposes the marginal variance of an outcome into components attributed to explanatory variables and a residual component. The resulting decomposit...

06617

Reposted by Aki Vehtari

Pierre-Simon Laplace @learnbayesstats.bsky.social · 31/08/2026

In this special episode of Learning Bayesian Statistics, @alex-andorra.bsky.social is joined by Andrew Gelman, Aki Vehtari, and Richard McElreath to discuss Bayesian Workflow. From simulation and hierarchical pooling to causal inference, there’s a lot to unpack 🎧 lnkd.in/geX2QkxV #Bayesian

learnbayesstats.com

Bayesian Workflow - Gelman, Vehtari & McElreath

Andrew Gelman, Aki Vehtari, and Richard McElreath discuss their new book on Bayesian workflow, reverse Bayes and hierarchical pooling.

13510

Reposted by Aki Vehtari

Osvaldo Martin @aloctavodia.bsky.social · 31/08/2026

We need reviewers for our submission to the Journal of Open Source Education. Our resource focuses on MCMC convergence, model checking, comparison, etc all powered by ArviZ. This is the issue where you can volunteer: github.com/openjournals...

github.com

[PRE REVIEW]: Exploratory Analysis of Bayesian Models · Issue #289 · openjournals/jose-reviews

Submitting author: @aloctavodia (Osvaldo Martin) Repository: https://github.com/arviz-devs/EABM Branch with paper.md (empty if default branch): Version: v0.2.0 Editor: Pending Reviewers: Pending Ma...

233

Reposted by Aki Vehtari

Jarno Vanhatalo @jarnovanhatalo.bsky.social · 31/08/2026

In our latest study we show that Baltic Sea pelagic fish communities have changed their distribution over the past 17 years, consistent with projected climate-driven salinity reductions 👇 Community-level modelling of pelagic fish assemblages... academic.oup.com/icesjms/arti... 1/3

academic.oup.com

Community-level modelling of pelagic fish assemblages in the Baltic Sea reveals temporal shifts and effects of environmental drivers

Abstract. The Baltic Sea is characterised by strong environmental gradients that shape its biodiversity and community structure. Despite a well-established

193

Aki Vehtari @avehtari.bsky.social · 26/08/2026

bayesplot R package 1.16.0 now in CRAN! This release has quite a lot of new functionality and other improvements. Below are some highlights. For more detailed release notes that cover additional improvements and bug fixes see Changelog mc-stan.org/bayesplot/ne...

mc-stan.org

Changelog

15012

Reposted by Aki Vehtari

Maximilian Scholz @scholzmx.bsky.social · 23/08/2026

Watched Bob Carpenter's StanCon talk and decided the Stan stack deserved some love. PRs incoming to stan-math, stanc3, bridgestan + walnutpie: −15% walltime per gradient, a silent autodiff blowup on degenerate eigenvalues (ESS 29→411 when fixed), and a small pile of bugs 👀 :3 #rstats #stan

2101

Aki Vehtari @avehtari.bsky.social · 21/08/2026

Given all the discussions about Laplace and Jacobians at StanCon, reminder also about my case study with simple visual illustrations users.aalto.fi/~ave/casestu...

users.aalto.fi

Laplace method and Jacobian of parameter transformation – Aki Vehtari

0174

Reposted by Aki Vehtari

Victor Van der Meersch @vvandermeersch.bsky.social · 20/08/2026

Had a wonderful time at StanCon in Uppsala! Glad to connect with such a thriving Bayesian community. Many thanks to the organizers for putting together this cool event

092

Reposted by Aki Vehtari

Matthew Kay @mjskay.com · 20/08/2026

Given all the discussions about Jacobians at #StanCon today, the @mc-stan.org audience would appreciate that {ggdist} automatically applies the Jacobian adjustment when visualizing densities on transformed scales -- I should have demoed it in my talk! Sometimes useful when looking at priors...

A lognormal distribution on the original scale and then on a log scale, showing how its shape is transformed (and is Normal on the log scale).

A normal distribution on the original scale and then on an inverse-logit scale, showing how its shape is transformed.

0184

Reposted by Aki Vehtari

Charles Margossian @charlesm993.bsky.social · 21/08/2026

As of v2.39, Stan (@mc-stan.org) provides an embedded Laplace approximation for fitting latent Gaussian models. Thanks to the tremendous effort by Steve Bronder, @avehtari.bsky.social, @brianward.dev, and many others. 📽️ www.youtube.com/watch?v=DDTP... 🧵 1/

youtube.com

Charles Margossian - Embedded Laplace Approximation in Stan. StanCon 2026

YouTube video by Stan

182

Aki Vehtari @avehtari.bsky.social · 21/08/2026

StanCon 2026 tutorial "Bayesian model diagnostics: Workflows and software tools" by Noa Kallioinen, @teemusailynoja.bsky.social and @aloctavodia.bsky.social is available online n-kall.github.io/stancon2026/

n-kall.github.io

Bayesian model diagnostics: Workflows and software tools

0266

Aki Vehtari @avehtari.bsky.social · 21/08/2026

ELLIS Institute Finland PostDoc and doctoral student call is open (DL September 21, 2026). You can apply also to work with me on computational methods for Bayesian workflows www.ellisinstitute.fi/postdoc-and-...

ellisinstitute.fi

Postdoc and doctoral student positions at ELLIS Institute Finland | ELLIS Institute Finland

Call for postdocs and doctoral students in artificial intelligence and machine learning

093

Aki Vehtari @avehtari.bsky.social · 20/08/2026

I forgot to post this yesterday. To counterbalance all sitting in a lecture hall at StanCon, we played football with a lot of divergent passes and shots

Photo of people on football court. One player is wearing StanCon 2026 Uppsala t-shirt

0140

Aki Vehtari @avehtari.bsky.social · 20/08/2026

StanCon 2026 has been awesome! The talks are online, but it's been great to also have many discussions in person!

0171

Reposted by Aki Vehtari

Mattias Villani @matvil.bsky.social · 20/08/2026

The StanCon2026 conference on probabilistic programming for Bayes in Uppsala is over. Very positive experience with interesting talks, a large poster session and social activities. And some HMC updates to modernize the Bayesian Songbook! All talks are on YouTube: youtube.com/playlist?lis...

youtube.com

StanCon 2026 Uppsala - YouTube

052

Reposted by Aki Vehtari

Sean Pinkney @spinkney.bsky.social · 20/08/2026

Stancon26 talk about a surprising discovery I had that Bayesian hierarchical models can be made equivalent in distribution to frequentist estimates, giving new insight into hierarchical models and faster inference. Upcoming paper with Nikolai Vetr youtu.be/lHXav9KIEIU?...

youtu.be

Sean Pinkney - Population Effects in Hierarchical Models using Sum-to-Zero Constraints. StanCon 2026

YouTube video by Stan

0235

Aki Vehtari @avehtari.bsky.social · 19/08/2026

Two StanCon 2026 talks related to papers I've co-authored - Charles Margossian: Embedded Laplace Approximation in Stan www.youtube.com/watch?v=DDTP... - Anna Elisabeth Riha: To select or not to select - predictively consistent priors instead of model selection www.youtube.com/watch?v=9wqb...

youtube.com

Charles Margossian - Embedded Laplace Approximation in Stan. StanCon 2026

YouTube video by Stan

0120

Aki Vehtari @avehtari.bsky.social · 19/08/2026

Revised version of "Bridge sampling diagnostics" with Giorgio Micaletto arxiv.org/abs/2508.14487 - extended experiments with reference values - better diagnostic for estimate reliability - new hybrid score-matching proposal significantly improving the stability of the estimator 1/4

Abstract:
In Bayesian statistics, the marginal likelihood is used for model selection and averaging, yet it is often challenging to compute accurately for complex models. Approaches such as bridge sampling, while effective, suffer from high variance when the proposal distribution overlaps poorly with the target posterior. To quantify this variance, we present a closed-form Monte Carlo standard error (MCSE) estimator for bridge sampling, extending classical variance approximations with a multi-chain effective-sample-size correction for autocorrelated MCMC draws and an exact log-scale variance. We show that the MCSE estimate itself is structurally capped at about 1.05, so values near this cap signal saturation rather than precision, and our calibration experiments show that the MCSE can be trusted when it is below 0.3. Furthermore, we introduce a hybrid score-matching proposal that regularizes the sample covariance using the local posterior geometry, significantly improving the stability of the estimator, and we demonstrate the efficacy of these methods using increasingly difficult simulated posteriors and real posteriors from the posteriordb database.

161

Reposted by Aki Vehtari

MC Stan @mc-stan.org · 18/08/2026

There is now StanCon 2026 Uppsala playlist www.youtube.com/playlist?lis... At the moment talks by @mjskay.com, @aseyboldt.bsky.social, @paulbuerkner.com, @charlesm993.bsky.social are up, and we'll keep uploading more during the week

youtube.com

StanCon 2026 Uppsala - YouTube

23717

Aki Vehtari @avehtari.bsky.social · 18/08/2026

As an Aalto postdoc @mansmag.bsky.social helped organizing StanCon 2018 Helsinki, and now he is one of the main organizers of StanCon 2026 Uppsala!

A photo of two men with StanCon t-shirts for StanCon 2018 Helisnkin and StanCon 2026 Uppsala

0171

Aki Vehtari @avehtari.bsky.social · 18/08/2026

@mjskay.com starting the StanCon 2026 talks with "Systematic uncertainty visualization design (and its foundation in probabilistic models)" All talks are recorded and will be made available after the conference

Photo of a person in front of a lecture hall

0385

Aki Vehtari @avehtari.bsky.social · 17/08/2026

@florencebockting.bsky.social starting the StanCon 2026 tutorial: "Contributing to Stan: A Developer’s Guide to the Core, Interfaces, and First Contributions" to be followed later by @brianward.dev and @bronder.deals

1231

Aki Vehtari @avehtari.bsky.social · 13/08/2026

All four books I've co-authored are freely available online for non-commercial use: - Bayesian Workflow at avehtari.github.io/Bayesian-Wor... Links to other three books are in the quoted post 👇 (too many books to fit in one post!)

avehtari.github.io

Bayesian Workflow book: Website – Bayesian Workflow book

Website for the Bayesian Workflow book by Gelman, Vehtari, McElreath, et al. — case studies, code, and exercises in R and Stan.

8283133

Aki Vehtari @avehtari.bsky.social · 13/08/2026

BayesComp 2027 mirror in France!

040

Reposted by Aki Vehtari

MC Stan @mc-stan.org · 11/08/2026

Call for StanCon 2027+ If you think you would like to organize the next StanCon (after Uppsala), click the link and read more! discourse.mc-stan.org/t/call-for-s...

discourse.mc-stan.org

Call for StanCon 2027+

We are looking for volunteers to organize the next StanCons for 2027 (and beyond!) If you are interested in making a proposal (or even just discussing the possibility of making one), please consider ...

032

Reposted by Aki Vehtari

Significance magazine @significancemag.bsky.social · 30/07/2026

Known as the "saviour of mothers", Ignac Semmelweis pioneered handwashing in hospitals, saving countless lives. Can we apply modern statistical techniques to his own data? New and free to read: buff.ly/4NUSgLc

012

Reposted by Aki Vehtari

Matthijs Hollanders @matthollanders.bsky.social · 01/08/2026

Anyone ever use the geometric Poisson distribution? If I've implemented it correctly in Stan, it's giving me better fits than negative binomial according to PSIS-LOO-CV with a big count occupancy model.

2103

Reposted by Aki Vehtari

Julia M. Rohrer @dingdingpeng.the100.ci · 30/07/2026

New blog post! Odds ratios are...a bit weird. Here I explain two problems (interpretation, non-collapsibility) and provide some simple recommendations for improved reporting. www.the100.ci/2026/07/30/r...

the100.ci

Reviewer notes: Odds ratios are really odd

When psychological researchers investigate binary outcomes, they routinely report odds ratios to quantify the strength of an effect or an association. This is quite understandable given that psycholog...

2428779

Reposted by Aki Vehtari

Richard McElreath 🐈‍⬛ @rmcelreath.bsky.social · 29/07/2026

For examples of the priorsense package in use, see e.g. chapter 17 case study ("sleep study") from the Bayesian Workflow website: avehtari.github.io/Bayesian-Wor...

figure 10 from the linked page, showing prior sensitivity checks for three parameters

0477

Aki Vehtari @avehtari.bsky.social · 29/07/2026

priorsense R package has now JOSS paper "priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R" you can cite, too doi.org/10.21105/jos... with Noa Kallioinen, Topi Paananen, and @paulbuerkner.com priorsense achieved also a gold badge from @ropensci.org review!

doi.org

priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R

Kallioinen et al., (2026). priorsense: Efficient prior and likelihood sensitivity checks for Bayesian models in R. Journal of Open Source Software, 11(123), 11036, https://doi.org/10.21105/joss.11036

16816