Sam Power
@spmontecarlo.bsky.social
Lecturer in Maths & Stats at Bristol. Interested in probabilistic + numerical computation, statistical modelling + inference. (he / him).
Homepage: https://sites.google.com/view/sp-monte-carlo
Seminar: https://sites.google.com/view/monte-carlo-semina
Homepage: https://sites.google.com/view/sp-monte-carlo
Seminar: https://sites.google.com/view/monte-carlo-semina
Eclectic research interests
February 4, 2026 at 12:27 AM
Eclectic research interests
How do I get Diaconis-Holmes-Neal as my loading screen?
February 3, 2026 at 5:56 PM
How do I get Diaconis-Holmes-Neal as my loading screen?
My most recent attempt at visualising what spectral / L2 convergence of Markov chains looks like. Not fully satisfied with the outcome (would somehow be better if I could relate these waves to the distribution of X_t over time), but I suppose that it does illustrate the global character a bit.
February 3, 2026 at 5:52 PM
My most recent attempt at visualising what spectral / L2 convergence of Markov chains looks like. Not fully satisfied with the outcome (would somehow be better if I could relate these waves to the distribution of X_t over time), but I suppose that it does illustrate the global character a bit.
Reposted by Sam Power
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....
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....
January 29, 2026 at 12:38 PM
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....
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....
Reposted by Sam Power
New open source: cuthbert 🐛
State space models with all the hotness: (temporally) parallelisable, JAX, Kalman, SMC
State space models with all the hotness: (temporally) parallelisable, JAX, Kalman, SMC
January 30, 2026 at 4:26 PM
New open source: cuthbert 🐛
State space models with all the hotness: (temporally) parallelisable, JAX, Kalman, SMC
State space models with all the hotness: (temporally) parallelisable, JAX, Kalman, SMC
At a glance, these look decent, even in terms of being quite suitable for analysts / probabilists (as opposed to just number theorists):
people.math.ethz.ch/~kowalski/pr...
people.math.ethz.ch/~kowalski/pr...
February 1, 2026 at 2:44 PM
At a glance, these look decent, even in terms of being quite suitable for analysts / probabilists (as opposed to just number theorists):
people.math.ethz.ch/~kowalski/pr...
people.math.ethz.ch/~kowalski/pr...
Some notes on the Central Limit Theorem (following my revisiting of some of Devroye's book on Random Variate Generation):
Often, in handling random variables for which 'direct' simulation is challenging, we try to 'reduce' to sampling of something tractable, and then post-process / transform / etc.
Often, in handling random variables for which 'direct' simulation is challenging, we try to 'reduce' to sampling of something tractable, and then post-process / transform / etc.
February 1, 2026 at 1:04 PM
Some notes on the Central Limit Theorem (following my revisiting of some of Devroye's book on Random Variate Generation):
Often, in handling random variables for which 'direct' simulation is challenging, we try to 'reduce' to sampling of something tractable, and then post-process / transform / etc.
Often, in handling random variables for which 'direct' simulation is challenging, we try to 'reduce' to sampling of something tractable, and then post-process / transform / etc.
some hits from recent reading:
arxiv.org/abs/2511.08111
'On the Kantorovich contraction of Markov semigroups'
- Pierre Del Moral, Mathieu Gerber
www.arxiv.org/abs/2601.19710
'On randomized step sizes in Metropolis-Hastings algorithms'
- Sebastiano Grazzi, Samuel Livingstone, Lionel Riou-Durand
arxiv.org/abs/2511.08111
'On the Kantorovich contraction of Markov semigroups'
- Pierre Del Moral, Mathieu Gerber
www.arxiv.org/abs/2601.19710
'On randomized step sizes in Metropolis-Hastings algorithms'
- Sebastiano Grazzi, Samuel Livingstone, Lionel Riou-Durand
On the Kantorovich contraction of Markov semigroups
This paper develops a novel operator theoretic framework to study the contraction properties of Markov semigroups with respect to a general class of Kantorovich semi-distances, which notably includes ...
arxiv.org
January 30, 2026 at 9:04 PM
some hits from recent reading:
arxiv.org/abs/2511.08111
'On the Kantorovich contraction of Markov semigroups'
- Pierre Del Moral, Mathieu Gerber
www.arxiv.org/abs/2601.19710
'On randomized step sizes in Metropolis-Hastings algorithms'
- Sebastiano Grazzi, Samuel Livingstone, Lionel Riou-Durand
arxiv.org/abs/2511.08111
'On the Kantorovich contraction of Markov semigroups'
- Pierre Del Moral, Mathieu Gerber
www.arxiv.org/abs/2601.19710
'On randomized step sizes in Metropolis-Hastings algorithms'
- Sebastiano Grazzi, Samuel Livingstone, Lionel Riou-Durand
Exciting stuff: next week at the Online Monte Carlo Seminar, we will finally have a bit of Randomised Numerical Linear Algebra, courtesy of Ethan Epperly (Berkeley). See our website (sites.google.com/view/monte-c...) for details about the talk, taking place next Tuesday (3 February).
January 30, 2026 at 6:36 PM
Exciting stuff: next week at the Online Monte Carlo Seminar, we will finally have a bit of Randomised Numerical Linear Algebra, courtesy of Ethan Epperly (Berkeley). See our website (sites.google.com/view/monte-c...) for details about the talk, taking place next Tuesday (3 February).
Currently watching a probability seminar about deposition models, KPZ, etc.; meanwhile, an audience member has Tetris open on their phone. There are many ways to capture an audience's attention, I suppose!
January 30, 2026 at 3:44 PM
Currently watching a probability seminar about deposition models, KPZ, etc.; meanwhile, an audience member has Tetris open on their phone. There are many ways to capture an audience's attention, I suppose!
powerful response on MO
January 26, 2026 at 4:49 PM
powerful response on MO
In the end, the answer: the lectures (from the first day) _were_ indeed recorded! Find them linked here: sites.google.com/view/lpd-tnn....
January 26, 2026 at 11:15 AM
In the end, the answer: the lectures (from the first day) _were_ indeed recorded! Find them linked here: sites.google.com/view/lpd-tnn....
Itching to have the space in my calendar to process this in detail:
arxiv.org/abs/2601.12633
'New Trends in the Stability of Sinkhorn Semigroups'
- Pierre Del Moral, Ajay Jasra
arxiv.org/abs/2601.12633
'New Trends in the Stability of Sinkhorn Semigroups'
- Pierre Del Moral, Ajay Jasra
January 22, 2026 at 8:18 PM
Itching to have the space in my calendar to process this in detail:
arxiv.org/abs/2601.12633
'New Trends in the Stability of Sinkhorn Semigroups'
- Pierre Del Moral, Ajay Jasra
arxiv.org/abs/2601.12633
'New Trends in the Stability of Sinkhorn Semigroups'
- Pierre Del Moral, Ajay Jasra
Kind of funny to read this and have the reaction that "that's not algebra, that's statistical physics".
January 21, 2026 at 9:39 PM
Kind of funny to read this and have the reaction that "that's not algebra, that's statistical physics".
Recent downloads: "Distribution-uniform anytime-valid inference and the Robbins-Siegmund distributions" (arxiv.org/abs/2311.033...).
January 20, 2026 at 6:16 PM
Recent downloads: "Distribution-uniform anytime-valid inference and the Robbins-Siegmund distributions" (arxiv.org/abs/2311.033...).
idle thought: if one thinks of the h-index as an estimator, then what 'population-level' quantity (if any) is it (or some simple transformation thereof) an estimator of?
January 19, 2026 at 10:09 AM
idle thought: if one thinks of the h-index as an estimator, then what 'population-level' quantity (if any) is it (or some simple transformation thereof) an estimator of?
Lovely new work from my friend (and former office mate!) Sven:
arxiv.org/abs/2601.09007
'Global polynomial-time estimation in statistical nonlinear inverse problems via generalized stability'
- Sven Wang
arxiv.org/abs/2601.09007
'Global polynomial-time estimation in statistical nonlinear inverse problems via generalized stability'
- Sven Wang
Global polynomial-time estimation in statistical nonlinear inverse problems via generalized stability
Non-linear statistical inverse problems pose major challenges both for statistical analysis and computation. Likelihood-based estimators typically lead to non-convex and possibly multimodal optimizati...
arxiv.org
January 15, 2026 at 8:04 PM
Lovely new work from my friend (and former office mate!) Sven:
arxiv.org/abs/2601.09007
'Global polynomial-time estimation in statistical nonlinear inverse problems via generalized stability'
- Sven Wang
arxiv.org/abs/2601.09007
'Global polynomial-time estimation in statistical nonlinear inverse problems via generalized stability'
- Sven Wang
Have seen this news (cims.nyu.edu/dynamic/news...) about NYU professor Robert Kohn. As good a time as any to share a piece of his work which I found cool: when studying deterministic control problems, one can still end up with second-order PDEs for their solution: math.nyu.edu/~kohn/papers....
January 15, 2026 at 4:59 PM
Have seen this news (cims.nyu.edu/dynamic/news...) about NYU professor Robert Kohn. As good a time as any to share a piece of his work which I found cool: when studying deterministic control problems, one can still end up with second-order PDEs for their solution: math.nyu.edu/~kohn/papers....
Reposted by Sam Power
lol I have such strong opinions about this (like you should legally be allowed to dunk anyone who says a penalty is a lot prior in a vat of very very cold water)
January 14, 2026 at 3:40 PM
lol I have such strong opinions about this (like you should legally be allowed to dunk anyone who says a penalty is a lot prior in a vat of very very cold water)
And heading back from it now - great couple of days.
On my way to this now! Should be fun - always exciting to spend time at events outside of my own usual community, get a sense of what is capturing their attention, etc.
On behalf of some friends, let me quickly advertise an event taking place in London, January 12-13 2026 (sites.google.com/view/lpd-tnn), with an overall focus on 'Geometric methods in probability'. Registration is free but required, and closes on December 1 (i.e. next Monday) - exciting stuff!
January 13, 2026 at 9:19 PM
And heading back from it now - great couple of days.
commencing imminently!
Excited for the return of our seminar for another season!
Link with details: sites.google.com/view/monte-c...
Link with details: sites.google.com/view/monte-c...
January 13, 2026 at 4:31 PM
commencing imminently!
Very neat stuff:
arxiv.org/abs/2601.05444
'What Functions Does XGBoost Learn?'
- Dohyeong Ki, Adityanand Guntuboyina
arxiv.org/abs/2601.05444
'What Functions Does XGBoost Learn?'
- Dohyeong Ki, Adityanand Guntuboyina
January 12, 2026 at 11:32 AM
Very neat stuff:
arxiv.org/abs/2601.05444
'What Functions Does XGBoost Learn?'
- Dohyeong Ki, Adityanand Guntuboyina
arxiv.org/abs/2601.05444
'What Functions Does XGBoost Learn?'
- Dohyeong Ki, Adityanand Guntuboyina
On my way to this now! Should be fun - always exciting to spend time at events outside of my own usual community, get a sense of what is capturing their attention, etc.
On behalf of some friends, let me quickly advertise an event taking place in London, January 12-13 2026 (sites.google.com/view/lpd-tnn), with an overall focus on 'Geometric methods in probability'. Registration is free but required, and closes on December 1 (i.e. next Monday) - exciting stuff!
January 12, 2026 at 11:02 AM
On my way to this now! Should be fun - always exciting to spend time at events outside of my own usual community, get a sense of what is capturing their attention, etc.