Lightnews — Scholar-powered news

Reposted by Alex Thiery

Quanta Magazine @quantamagazine.bsky.social · Jul 7

In the world of sphere packing, there’s been debate about whether order or a dash of chaos will give the best results. A recent proof marks a win for order. www.quantamagazine.org/new-sphere-p...

6 23

Alex Thiery @alexxthiery.bsky.social · Jun 23

And a recent very well written review of NS:

"Nested sampling for physical scientists"

arxiv.org/abs/2205.15570

2

Alex Thiery @alexxthiery.bsky.social · Jun 23

Nested Sampling is extremely popular in some communities, and there are often claims that it helps mitigate "phase transition" issues that can often affect standard geometric "tempering" methods (although I do not understand that well enough yet...) It's great to see explicit connections with SMC!

1

Alex Thiery @alexxthiery.bsky.social · Jun 23

"Unbiased and Consistent Nested Sampling via Sequential Monte Carlo"

by Robert Salomone, Leah F. South, Christopher Drovandi, Dirk P. Kroese, Adam M. Johansen

arxiv.org/abs/1805.03924

Unbiased and Consistent Nested Sampling via Sequential Monte Carlo

We introduce a new class of sequential Monte Carlo methods which reformulates the essence of the nested sampling method of Skilling (2006) in terms of sequential Monte Carlo techniques. Two new algori...

arxiv.org

1

Alex Thiery @alexxthiery.bsky.social · Jun 23

"A simpler nested sampling identity"

Interesting blogpost on nested sampling & SMC by Nicolas Chopin

statisfaction-blog.github.io/posts/04-06-...

A simpler nested sampling identity – Statisfaction - I can’t get no

statisfaction-blog.github.io

1 7

Alex Thiery @alexxthiery.bsky.social · Jun 15

See you in 🇸🇬

2

Alex Thiery @alexxthiery.bsky.social · Jun 13

My bad, this wasn't clear. It's in the space of all probability densities

Alex Thiery @alexxthiery.bsky.social · Jun 13

Motivated by the reading of this nice article:
"Sequential Monte Carlo approximations of Wasserstein--Fisher--Rao gradient flows"
by Francesca R. Crucinio, Sahani Pathiraja
arxiv.org/abs/2506.05905

Sequential Monte Carlo approximations of Wasserstein--Fisher--Rao gradient flows

We consider the problem of sampling from a probability distribution $π$. It is well known that this can be written as an optimisation problem over the space of probability distribution in which we aim...

arxiv.org

1 9

Alex Thiery @alexxthiery.bsky.social · Jun 13

And here is how the geodesic path looks like (again under the Fisher-Rao metric)

1 6

Alex Thiery @alexxthiery.bsky.social · Jun 13

Here's how the gradient flow for minimizing KL(pi, target) looks under the Fisher-Rao metric. I thought some probability mass would be disappearing on the left and appearing on the right (i.e. teleportation), like a geodesic under the same metric, but I was very wrong... What's the right intuition?

4 6 23

Reposted by Alex Thiery

Christoph Dellago @chhdellago.bsky.social · Jun 7

Once you have tried symplectic integrators, you never go back.

6 3 46

Reposted by Alex Thiery

Sam Power @spmontecarlo.bsky.social · Jun 1

The full (?) program of talks etc. for BayesComp seems to be online now (bayescomp2025.sg#programme), and looks pretty exciting - I will need to set aside some time to carve out my own schedule!

3 26

Alex Thiery @alexxthiery.bsky.social · May 17

Once the prompt is public, I do not think it will provide much signal (but it could potentially slightly help some the papers make sure their writing style align well with the conference expectations)

1 1

Alex Thiery @alexxthiery.bsky.social · May 17

How to implement this in practice, make the "review" prompt public in advance?

1 1

Reposted by Alex Thiery

Gilles Louppe @glouppe.bsky.social · Apr 29

<proud advisor>
Hot off the arXiv! 🦬 "Appa: Bending Weather Dynamics with Latent Diffusion Models for Global Data Assimilation" 🌍 Appa is our novel 1.5B-parameter probabilistic weather model that unifies reanalysis, filtering, and forecasting in a single framework. A thread 🧵

2 15 50

Alex Thiery @alexxthiery.bsky.social · Apr 19

These sparse Gaussian Processes have been around longer than some grad students, but still fun to code! (and today was my first time coding one...)

1 8

Alex Thiery @alexxthiery.bsky.social · Apr 11

Today, re-reading a classic.. the 1953 paper that started it all

9

Alex Thiery @alexxthiery.bsky.social · Apr 1

Is it based on the last year's preprint by Huhtikuun Typerys?

1 2

Alex Thiery @alexxthiery.bsky.social · Apr 1

extracted from:
"Upper Bounds for the Connective Constant of Self-Avoiding Walks" by Sven Erick Alm
www.cambridge.org/core/journal...

Upper Bounds for the Connective Constant of Self-Avoiding Walks | Combinatorics, Probability and Computing | Cambridge Core

Upper Bounds for the Connective Constant of Self-Avoiding Walks - Volume 2 Issue 2

www.cambridge.org

2

Alex Thiery @alexxthiery.bsky.social · Apr 1

Cute way to upper bound the connective constant of Z^d. For some length L, enumerate {w_1, w_2, ... , w_N} the Self-Avoiding-Walks of size L. An upper bound is given by the largest eigenvalue of the NxN matrix where M_{i,j}=1 iff there is a SAW of size (L+1) that starts with w_i and ends with w_j.

1 1

Alex Thiery @alexxthiery.bsky.social · Apr 1

Ah, but this paper seems to be confident that the conjecture is wrong, based on extensive simulations for estimating the connective constant up to 12 decimals (at which point there is a departure from the conjectured value). Still open though 😅
arxiv.org/pdf/1607.02984

arxiv.org

1

Alex Thiery @alexxthiery.bsky.social · Apr 1

Conjecture dates from 1992:
"Algebraic Techniques for Enumerating Self-Avoiding Walks on the Square Lattice"
arxiv.org/abs/hep-lat/...

"While we consider it would be fortuitous if this were the true value of the critical point, it nevertheless provides a useful mnemonic" 🙂

Algebraic Techniques for Enumerating Self-Avoiding Walks on the Square Lattice

We describe a new algebraic technique for enumerating self-avoiding walks on the rectangular lattice. The computational complexity of enumerating walks of $N$ steps is of order $3^{N/4}$ times a polyn...

arxiv.org

1 2

Alex Thiery @alexxthiery.bsky.social · Apr 1

Approximating N(L), the number of Self-Avoiding-Walks in Z^2 of length L, is an assignment in my Simulation course this year. The connective constant is:

C = \lim N(L)^1/L ~ 2.638..

Still open-problem to this day: is it true that 1/C equals the zero of the polynomial P(x)=581*x^4 + 7*x^2 - 13 😱

1 5

Alex Thiery @alexxthiery.bsky.social · Mar 30

That's interesting that it seems like very little is known about the asymptotic of the second largest increasing subsequence (and no fast method to compute it)

3

Alex Thiery @alexxthiery.bsky.social · Mar 30

This fast way of finding the LIS is neat! Just tried to reproduce your nice plot without leaving the phone 😊
chatgpt.com/share/67e8ec...

1 1