Post on solving problems with LLMs with some #Rstats examples: www.visruth.com/blog/2026101...
Solving Problems with LLMs
Visruth Srimath Kandali's Website
visruth.com
Visruth Srimath Kandali
@visruth.com
Cal Poly Statistics student broadly interested in Bayesian statistics, computational methods, and modeling. Member of the Stan development team and R-multiverse moderator.
Post on solving problems with LLMs with some #Rstats examples: www.visruth.com/blog/2026101...
Solving Problems with LLMs
Visruth Srimath Kandali's Website
visruth.com
CmdStanR 1.0 release candidate is now available! There are a lot of changes! Please read more and help testing discourse.mc-stan.org/t/cmdstanr-1... (CmdStanR is an R interface for using Stan via CmdStan)
CmdStanR 1.0 release candidate is now available
I’m happy to announce that a release candidate for CmdStanR 1.0 is ready and available for testing! This will be by far the biggest release since the initial release of the package. One of the main ch...
discourse.mc-stan.org
Louder for the people in the back: Jev can’t be calibrated www.alexmolas.com/2026/09/23/j... Friends don’t let friends lie about the accessibility of the DGD even in principle
Jev can't be calibrated
Jev is a useful zero-shot classifier, but its probabilities can't be calibrated for your data. Calibration depends on your data distribution, which Jev never sees, so treat its outputs as scores and r...
alexmolas.com
New beta of #Rstats rig: LOTS of EXCITING stuff: - new: web site: rig.r-lib.org - new: user-mode R installation: rig.r-lib.org/admin-vs-use... - new: manage R packages (CRAN + git): rig.r-lib.org/pkg-guide.html - new: manage R projects and virtual environments: rig.r-lib.org/proj-guide.h...
rig – the R Installation Manager – rig
rig.r-lib.org
`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...
`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...
discourse.mc-stan.org
Just released a pkgdown template #Rstats package driven by a single _brand.yml file. Has some scaffolding to make using it nice in pkgdown and with #Quarto. www.visruth.com/blog/2026080...
Brand template repo
Visruth Srimath Kandali's Website
visruth.com
I have been tinkering with how to quickly modify a Quarto theme, and I think I have a decently fast solution. Introducing the Quarto Theme Tweaker Run the `Open Theme Tweaker` command in an active qmd file. Modify the settings to your liking. Save the results. github.com/EmilHvitfeld... #quarto
Just released bootstrapper 1.1.0, now with support for setting up touchstone and cleaner support for using air/jarl (especially in an existing project). bootstrapper is a #rstats package designed to quickstart new packages, following some practices I like. Read more: github.com/VisruthSK/bo...
GitHub - VisruthSK/bootstrapper: Opinionated R package quickstart
Opinionated R package quickstart. Contribute to VisruthSK/bootstrapper development by creating an account on GitHub.
github.com
The 2026 Rousseeuw Prize for Statistics has been awarded to the R Project. #RStats www.rousseeuwprize.org/2026
The Rousseeuw Prize for Statistics
The Rousseeuw Prize for Statistics is a biennial prize to celebrate outstanding contributions to statistics research.
rousseeuwprize.org
🚨 @rmcelreath.bsky.social @statmodeling.bsky.social & @avehtari.bsky.social are coming on the show mid-June to discuss their new book, Bayesian Workflow! ONE listener gets to bring a real Bayesian problem onto the recording and have the three of them work through it live. Here's how to enter 🧵
Wow the new codex usage limits really sting. Or maybe 5.5 is just token inefficient? Either way I'm hitting my 5hr limit consistently now, even only on medium. School pays though, so I guess I can't really complain.
New post on AI, self-preserving compassion, and how none of us are immune to it. open.substack.com/pub/jessicah...
More details about the Bayesian Workflow book and case studies now available on the book web site avehtari.github.io/Bayesian-Wor... (but you still need to wait a bit for the book)
Bayesian Workflow book: Website – Bayesian Workflow book
avehtari.github.io
I would have preferred to have the "draw the rest of the owl" meme on the cover, but this will do. Seems like it is on schedule, and we'll leave some typos so you know we didn't write it with AI.
By now it is obvious that AI has completely solved statistics and machine learning. So as the next step in the tidymodels journey we are moving into game development! Introducing our first game! emilhvitfeldt.github.io/tidymodels-d...
v1 of `bootstrapper`, an opinionated #rstats package to help you start building modern R packages. Wraps a number of usethis funcs to simplify starting new packages. Adds some common GHA, dependabot, air + jarl (precommit hook and GHA), and a concise AGENTS.md which I'm rather liking.
Opinionated R Package Quickstart
Bootstrap new R packages with strongly opinionated defaults. Creates the package scaffold, wires up tests, README and NEWS, configures GitHub Actions and Dependabot, adds Air/Jarl setup, and drop in a...
bootstrapper.visruth.com
Fantastic post by Colin Raffel, "We Are Over-Indexing on Paper Acceptance," drafted in May 2021 (!) but only posted now. The more things change.. Last sentence: "If you want to judge a researcher’s quality, the only meaningful way is to read their papers and judge for yourself."
Repurposed a bit of an old post on aesthetics and hastily blended some other thoughts. tl;dr: I want science fiction artificial intelligences like Olivaw et al., and I want to reserve the term AI for that kind of intelligences, not for LLMs. www.visruth.com/writings/ai/
What's an AI?
Visruth Srimath Kandali's Website
visruth.com
Its big problem with openclaw is that its just a terrible product that does nothing of value and a big company buys it and ignores real tools that do real work
Got the big Calvin and Hobbes collection and it is bananas that Watterson dropped one of the best strips of all time within a month of starting the comic
First release of stanflow! v0.1.0 was a few days later than I'd like, but its up now. Stanflow is a metapackage a la tidyverse for a Stan Bayesian workflow--see the README for more details/features! Feedback/issues/PRs are always appreciated. #rstats #bayes
GitHub - VisruthSK/stanflow: R Package for a Mildly Opinionated Stan Bayesian Workflow
R Package for a Mildly Opinionated Stan Bayesian Workflow - VisruthSK/stanflow
github.com
They found that much of LLM “reasoning” doesn’t come from RL training; it comes from how you sample the model. Paper: Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening ( www.arxiv.org/abs/2601.21590 )
All the material for my Bayesian Data Analysis course is available online, including the lectures, which we re-recorded this fall (some of them by @aloctavodia.bsky.social and Noa Kallioinen while I was on vacation). The video links are listed in the schedule at avehtari.github.io/BDA_course_A...
Bayesian Data Analysis course - Aalto 2025 – Bayesian Data Analysis course
avehtari.github.io
I'm teaching Statistical Rethinking again starting Jan 2026. This time with live lectures, divided into Beginner and Experienced sections. Will be a lot more work for me, but I hope much better for students. I will record lectures & all will be found at this link: github.com/rmcelreath/s...
Huge thanks to @posit.co and hosts @mchow.com and @wesmckinney.com for having me on The Test Set!!! posit.co/thetestset/e... My favorite part was learning that I've been pronouncing "pandas" correctly all along. I've never felt more vindicated. And y'all are wrong about candy corn. 😜
Episode 11 : Kelly Bodwin — Quarto hacks, AI in the classroom, and why R should stay weird - Posit
posit.co
Been really confused by social media for a while but I think I've figured it out: everyone except me has an explosive collar and it will detonate if they don't immediately contribute to the same discourse that everyone hates
"By the way, to unlock the full functionality of all Apps, enable Gemini Apps Activity." Gemini complains from time to time about me not enabling apps activity in the webapp--I added an instruction or wtv Google calls it to forbid it for doing so, but it persists. It really ticks me off.
Plugging my two favorite #python tools: uv and @marimo.io: www.visruth.com/blog/2025112... #rstats #julia too, because I always manage to mention R and Julia :)
`uv` & `marimo`
Visruth Srimath Kandali's Website
visruth.com
It’s easy to talk about ‘large AI data centers’ and still underestimate the scale. Our Frontier Data Centers database shows that some upcoming campuses will cover a substantial portion of Manhattan. Meta's Hyperion data center will be nearly four times the size of Central Park.