Archive for metro station

a journal of the imperial centuries

Posted in Books, Kids, Mountains, pictures, Running, Travel, University life, Wines with tags , , , , , , , , , , , , , , , , , , , , , , , , , on February 9, 2026 by xi'an

Over my trip to Rome, I managed to finish Titanium Noir, a science-fiction novel by Nick Harkaway, about an old-school detective trying to solve a murder in a society where a happy fews can become Titans, enhanced humans (aka transhumans) with larger bodies and higher life expectancy, except when they get killed of course. This is a rather enjoyable book, even though the science-fiction part does not play so well, since almost everything sounds like it was the 50’s, the 1950’s! Even the dialogues, which are somewhat outdated (no one uses gumshoe anymore!). But apart from that, the detective work is enough of a page turner and the final twist not completely predictable. Enjoyable maybe not to the point of continuing the series, since the twist cannot happens twice! Also read the very short novella Human Resources by Adrian Tchaikovsky, hugely if shortly disappointing!

Cooked winter standards (at home) like butternut soup and pulled pork. And pissaladière. Had two meals made of tonnarelli pasta, while 25H in Rome, a square variant to the spaghetti. One before strolling to the Fori Romani to watch the most recent excavations. And to admire the new Colosseo-Fori Imperiali metro station (with reconstituted household wells! And exhibits of objects found at their bottom).

Unrelated, but I watched MIU 404, a rather silly police series set in Tokyo, whose main appeals resides in exposing some issues in the modern Japanese society, like stalking or the status of migrants. But don’t expect realistic resolutions of the crime(s) motivating each episode…

gradient flow for projected Langevin dynamics

Posted in Books, Statistics, University life with tags , , , , , , , , , , , , , , on April 7, 2025 by xi'an

Daniel Lacker (Columbia U) gave a talk at the probability seminar of Paris Dauphine this week which I happened to attend by happenstance, on a recent paper, Projected Langevin dynamics and a gradient flow for entropic optimal transport, written with Giovanni Conforti and, Soumik Pal. The talk was quite progressive and I hence could follow most of it. The core idea is in studying Langevin-type diffusion dynamics that sample from an entropy-regularized optimal transport, i.e. looking for an optimal distribution (in the sense of achieving entropy minimisation problem within a Wasserstein space, with regularisation) obtained via a gradient flow equation (as eg in variational inference) that couples two SDEs that are recentred by conditional expectation terms. Expectations in the equations are estimated by a Nadaraya-Watson estimate in optimal transport problem (reminding me of SMC), with no theoretical derivation of an optimal bandwidth, and they achieve quantitive bounds on the convergence, namely for exponential convergence, energy decay and new logarithmic Sobolev inequalities. From the talk and a quick glance at the paper, it is unclear to me there are direct algorithmic consequences, since the SDEs need be discretised, while the expectation approximations are costly, being repeated at each iteration of the discretised SDE.