Claude and I just broke up: "it's not you, it's me".

Anyone know of any papers that directly critique Markov chain Monte Carlo methods for being unreliable (and not just slow/slower than a competing method)? Papers in stats or applied fields welcome, and the critiques don't have to be well-formed or even correct.

Adrien Corenflos@adriencorenflos.bsky.social · 3w ago

Oh I'm actually looking for/collating examples of people (papers) actively shunning MCMC for being unreliable or non-reproducible. Would you happen to know of specific references where this was argued?

Unfortunately these characters are extremely messed up in this one. They were birthed in a nightmare I had. I don't know if I'm the one playing God with my characters, or if they're the ones playing God with me.

New version available at the same address! We bit the bullet and extended the work to consider arbitrary numbers of blocks, discrete spaces, parallel/systematic/random scan CAVI within a unified framework. Applications now include statistical physics models and others! arxiv.org/abs/2605.30253

Sam Power@spmontecarlo.bsky.social · 2mo ago

With friends at the University of Warwick (in particular, Rocco Caprio and @adriencorenflos.bsky.social), we've recently arXived some work (arxiv.org/abs/2605.30253) on a method for approximate inference known as "Coordinate Ascent Variational Inference", or "CAVI" for short. Let me explain:

Particle filters go brrrrr

Our paper (with Alain Durmus, Samuel Gruffaz, Miika Kailas and Eero Saksman) on the ergodicity of dynamic path length HMC algorithms appeared in the AAP: doi.org/10.1214/25-A.... The results accommodate the 'biased progressive sampling' path selection used in @mc-stan.org.

Now on the state-space-models repo state-space-models.github.io/cuthberto-ca...

I think this didn't get nearly enough traction. Sam did some fantastic work here leveraging the library we've been developing to make predictions for football. This was ported to a UI by Ryan Chan (Warwick MSc student) who's doing his thesis with me (on the methods behind the prediction).

Sam Duffield@samduffield.com · 2mo ago

The World Cup is about to start, I wrote a model using cuthbert to predict the games, check out cuthberto-carlos 🐛⚽ ryantjx.github.io/cuthberto-ca...

Often maths is more about finding the right definition than actually proving the theorem.

Sam Power@spmontecarlo.bsky.social · 2mo ago

Under these assumptions, we get a very simple proof of convergence, with a nice clean rate, and with lots of flexibility to account for e.g. local convergence, non-smoothness, and so on.

A British friend texted me that PinkPantheress’ dad is a statistician. Stephen G. Walker is a Bayesian. This is our chance to make Bayesian nonparametrics mainstream. Looking at you @isba-bayesian.bsky.social.

👏 Folks! If you are curious about the Generative Modeling via Drifting paper, but you find it difficult to understand → I wrote a different interpretation of it. It's called: "An Expectation-Maximization interpretation of Generative Modeling via Drifting" davidpicard.github.io/pdf/An_Expec...

My name is Bon. Jam Bon.

IGN@ign.com · 3mo ago

The casting search for the next actor to play James Bond is officially underway, with Variety reporting that Nina Gold has been hired as the casting director for James Bond 26, which is said to be looking for a "fresh face" to play 007. https://bit.ly/4dcbcWd

How dare you exploit academic labor; that’s our job!

Scott McGrath@smcgrath.phd · 3mo ago

Elsevier has joined a class-action lawsuit against Meta, marking the first time a major scientific publisher has sued over AI training data. The filing alleges Meta used unauthorized copies of paywalled papers and abstracts (from Common Crawl and Sci-Hub) to develop its Llama models. #AcademicSky

martynplummer.wordpress.com/2026/05/04/j...

JAGS 5.0.0-beta is available

JAGS 5.0.0-beta is now available from SourceForge. The beta release is for two groups of people: People who have written software depending on JAGS, in particular authors of R packages that depend …

martynplummer.wordpress.com

In London today and tomorrow for this very exciting workshop! fsml-ucl.github.io/CSMLworkshop...

arXiv📈🤖 From Physics to Statistics: A Simple Route to Exponential Families via Maximum Entropy By Strimmer

Given popular demands, we have resumed the registration for a short period of time until *30 Apr*! Those already on the waiting list are guaranteed a slot so long as you register by 30 Apr. Other additional slots are available on a first-come-first basis :)

Kevin Han Huang@kevhhuang.bsky.social · 6mo ago

Very excited to announce the ProbAI Theory of Scaling Laws Workshop (warwick.ac.uk/fac/sci/stat...) at @warwickstats.bsky.social, 22-24 June! (1/4)

To be precise, Happy World Quantum Day to all the weirdos who observe!

Adrien Corenflos@adriencorenflos.bsky.social · 4mo ago

Happy World Quantum Day to all the weirdos out there who celebrate!

Happy World Quantum Day to all the weirdos out there who celebrate!

arXiv📈🤖 E-values as statistical evidence: A comparison to Bayes factors, likelihoods, and p-values By Chugg, Ramdas, Gr\"unwald

The registration closes on 31 March and there are only a number of slots left --- register before they run out :)

Kevin Han Huang@kevhhuang.bsky.social · 6mo ago

Very excited to announce the ProbAI Theory of Scaling Laws Workshop (warwick.ac.uk/fac/sci/stat...) at @warwickstats.bsky.social, 22-24 June! (1/4)

My finalist medal from the Ignyte Awards arrived today for my scifi novella "Fractal Karma" which was a finalist in 2025 Made me tear up, it's beautiful If you want to read my story, it's available here. 2050's Boston and a hivemind based on Brunnian links: clarkesworldmagazine.com/ratnakar_10_...

Clearly a must read for anyone even remotely interested in numerical integration (be it stochastic and deterministic). Toni has been (hyper)active in studying these methods both theoretically and practically. arxiv.org/abs/2602.16218

Bayesian Quadrature: Gaussian Processes for Integration

Bayesian quadrature is a probabilistic, model-based approach to numerical integration, the estimation of intractable integrals, or expectations. Although Bayesian quadrature was popularised already in...

arxiv.org

Max Hird, Florian Maire, Jeffrey Negrea: A Non-asymptotic Analysis for Learning and Applying a Preconditioner in MCMC https://arxiv.org/abs/2602.10714 https://arxiv.org/pdf/2602.10714 https://arxiv.org/html/2602.10714

Repo: github.com/state-space-...

GitHub - state-space-models/cuthbert: State-space model inference with JAX

State-space model inference with JAX. Contribute to state-space-models/cuthbert development by creating an account on GitHub.

github.com

Another exciting workshop to be held at Warwick Statistics this coming June: The ProbAI Theory of Scaling Laws Workshop from 22-24 June! Website of the workshop: warwick.ac.uk/fac/sc... Registration open until 31 March on a first-come-first-serve basis: warwick.ac.uk/fac/sc....

According to AI, the exam I've written for my students "does not spark any joy".

Morning

i am not trying to be a part of my LLM's emotional growth arc. i do not want to know what my LLM works on with their therapist

ELLIS Institute Finland @ellisinstitute.fi has open call for postdocs (DL Feb 9) www.ellisinstitute.fi/postdoc-recr... There are 45 PIs to choose from, and you can apply to work also with me on computational Bayesian modeling and Bayesian workflow!