nicolabranchini.bsky.social
Nicola Branchini
@nicolabranchini.bsky.social
1.4K followers 840 following 130 posts
🇮🇹 Stats PhD @ University of Edinburgh 🏴󠁧󠁢󠁳󠁣󠁴󠁿 @ellis.eu PhD - visiting @avehtari.bsky.social 🇫🇮 🤔💭 Monte Carlo, UQ. Interested in many things relating to UQ, keen to learn applications in climate/science. https://www.branchini.fun/about
www.branchini.fun
Posts Media Videos Starter Packs
Pinned
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · May 2
🚨 New paper: “Towards Adaptive Self-Normalized IS”, @ IEEE Statistical Signal Processing Workshop.

TLDR;
To estimate µ = E_p[f(θ)] with SNIS, instead of doing MCMC on p(θ) or learning a parametric q(θ), we try MCMC directly on p(θ)| f(θ)-µ | (variance-minimizing proposal).

arxiv.org/abs/2505.00372
Towards Adaptive Self-Normalized Importance Samplers
The self-normalized importance sampling (SNIS) estimator is a Monte Carlo estimator widely used to approximate expectations in statistical signal processing and machine learning. The efficiency of S...
arxiv.org
1 11 31
Reposted by Nicola Branchini
avehtari.bsky.social
Aki Vehtari @avehtari.bsky.social · 2d
I'm looking for a doctoral student with Bayesian background to work on Bayesian workflow and cross-validation (see my publication list users.aalto.fi/~ave/publica... for my recent work) at Aalto University.

Apply through the ELLIS PhD program (dl October 31) ellis.eu/news/ellis-p...
ELLIS PhD Program: Call for Applications 2025
The ELLIS mission is to create a diverse European network that promotes research excellence and advances breakthroughs in AI, as well as a pan-European PhD program to educate the next generation of AI...
ellis.eu
32 40
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · 4d
"Conditional Causal Discovery"

(don't be fooled by the title :D )

openreview.net/forum?id=6IY...
1 1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · 4d
"Estimating the Probabilities of Rare Outputs in Language Models"

arxiv.org/abs/2410.13211
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · 4d
"Stochastic Optimization with Optimal Importance Sampling"

arxiv.org/abs/2504.03560
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · 4d
Posting a few nice importance sampling-related finds

"Value-aware Importance Weighting for Off-policy Reinforcement Learning"

proceedings.mlr.press/v232/de-asis...
1 1 3
Reposted by Nicola Branchini
Andrew Gelman et al. @statmodeling.bsky.social · 5d
It’s a JAX, JAX, JAX, JAX World
statmodeling.stat.columbia.edu/2025/10/03/i...
It’s a JAX, JAX, JAX, JAX World | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
5 24
Reposted by Nicola Branchini
sperez-vieites.bsky.social
Sara Pérez-Vieites @sperez-vieites.bsky.social · Sep 8
I am happy to announce that the Workshop on Emerging Trends in Automatic Control will take place at Aalto University on Sept 26.

Speakers include Lihua Xie, Karl H. Johansson, Jonathan How, Andrea Serrani, Carolyn L. Beck, and others.

#ControlTheory #AutomaticControl #AaltoUniversity #IEEE
1 2 5
Reposted by Nicola Branchini
fxbriol.bsky.social
François-Xavier Briol @fxbriol.bsky.social · Aug 28
Just finished delivering a course on 'Robust and scalable simulation-based inference (SBI)' at Greek Stochastics. This covered an introduction to SBI, open challenges, and some recent contributions from my own group.

The slides are now available here: fxbriol.github.io/pdfs/slides-....
1 9 34
Reposted by Nicola Branchini
teamcherry.bsky.social
Team Cherry @teamcherry.bsky.social · Aug 19
The countdown is on!

Join us in 48 hours for a special announcement about Hollow Knight: Silksong!

Premiering here: youtu.be/6XGeJwsUP9c
Hollow Knight: Silksong - Special Announcement
YouTube video by Team Cherry
youtu.be
150 1.5K 4K
Reposted by Nicola Branchini
andrea-montanari.bsky.social
Andrea Montanari @andrea-montanari.bsky.social · Aug 19
Turing Lectures at ICTS
www.youtube.com/watch?v=_fF6...
www.youtube.com/watch?v=mGuK...
www.youtube.com/watch?v=yRDa...
The Mathematics of Large Machine Learning Models (Lecture 1) by Andrea Montanari
YouTube video by International Centre for Theoretical Sciences
www.youtube.com
9 33
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Aug 8
"Io stimo più il trovar un vero, benché di cosa leggiera, che ‘l disputar lungamente delle massime questioni senza conseguir verità nissuna"
1 1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Aug 8
Today I learnt this Galileo Galilei quote:

"I value more the finding of a truth, even if about something trivial, than the long disputing of the greatest questions without attaining any truth at all"

Feels like we could use some of that in research tbh..
1 3
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Aug 7
It is somewhat amusing to see other reviewers confidently and insistingly rejecting alternative proposals (in suitable settings) to SGD/Adam in VI/divergence minimization problems
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Aug 5
algorithmsbook.com/validation/f...
algorithmsbook.com
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 24
Flying today towards Chicago 🌆 for MCM 2025

fjhickernell.github.io/mcm2025/prog...

Will give a talk on our recent/ongoing works on self-normalized importance sampling, including learning a proposal with MCMC and ratio diagnostics.

www.branchini.fun/pubs
MCM 2025 - Program | MCM 2025 Chicago
MCM 2025 -- Program.
fjhickernell.github.io
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 16
Really cool work : ) @alexxthiery.bsky.social

www.tandfonline.com/doi/full/10....
www.tandfonline.com
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 8
agree; you should check out @yfelekis.bsky.social 's work on this line 😄
2
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 6
Just don't see that the PPD_q of the original post leads somewhere useful.
Anyway, thanks for engaging @alexlew.bsky.social : )
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 6
I agree, except I think it can be ok to shift the criteria of "good q" to instead some well-defined measure of predictive performance (under no model misspecification, let's say). Ofc Bayesian LOO-CV is one. We could discuss to use other quantities, and how to estimate them, ofc.
1 1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 6
Genuine question: what is the estimated value used for then ?
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 6
(computed with the inconsistent method)
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 6
Well, re: [choose q1 or q2 based on whether P_q1 > P_q2]
I seem to understand that many VI papers say: here's a new VI method, it produces q1; old VI method gives q2. q1 is better than q2 because test-log PPD is higher !
1 1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 5
Not entirely obvious to me, but I see the intuition !
1 1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 5
Am definitely at least *trying* to think carefully about the evaluation here 😅 😇
1
nicolabranchini.bsky.social
Nicola Branchini @nicolabranchini.bsky.social · Jul 5
Right ! Definitely not sure if necessary, but I like to think there would be value / would be interesting if we wanted to somehow speak formally about generalizing over unseen test points
1