Tootfinder

Zohran Mamdani - Seems like a rare ray of positivity!

Week 1 NFL player props, odds, predictions: Back Ja'Marr Chase over 83.5 yards for Week 1 NFL WR prop picks

https://www.cbssports.com/nfl/news/week-1-

83% in #newJersey

Now that we can put stories out there on #pixelfed I might start posting some on my trip, like a sort of diary.
#Fediverse
#blog

#EU5 working nice in #linux so far!
#proton

Charles Bierbauer, a former CNN correspondent and White House Correspondents' Association president from 1991 to 1992, has died at the age of 83 (Alison Main/CNN)
https://www.cnn.com/2025/08/31/politics/charles-bierbauer-correspondent-dies-83

Charles Bierbauer, former CNN correspondent, dies at 83
Bierbauer retired from CNN in 2001 after two decades covering news in Washington, DC, and around the world.

Dichte der #Windkraftanlagen je #Bundesland pro 10 km² in Deutschland mit Stand vom 04.10.2025.
Nur in Betrieb befindliche Anlagen berücksichtigt.
Die „Ausschließliche Wirtschaftszone“ (Offshore) ist ausgenommen.
👉 Zusatzlesestoff:

Dienstmededelung:
De pruimentaart staat in de oven
Ik herhaal:
De pruimentaart staat in de oven!
https://cooking.nytimes.com/recipes/3783-original-plum-torte?unlocked_article_code=1.jk8.7rf0.mSIJOO7HJ6Aw

Original Plum Torte Recipe (Gift Recipe)
The Times published Marian Burros’s recipe for Plum Torte every September from 1983 until 1989, when the editors determined that enough was enough. The recipe was to be printed for the last time that year. “To counter anticipated protests,” Ms. Burros wrote a few years later, “the recipe was printed in larger type than usual with a broken-line border around it to encourage clipping.” It didn’t help. The paper was flooded with angry letters. “The appearance of the recipe, like the…

#Wordle 1,539 3/6*
🟨⬜⬜⬜⬜ <1% of 213,416 (36)
⬜🟨🟩🟩⬜ 0 of 98 (2)
🟩🟩🟩🟩🟩
WordleBot
Skill 83/99
Luck 75/99
Such a fortuitous first word.
And my 2nd guess worked out but user error, reused an eliminated letter. Really not sure how that got by me.

The new PixelFed update looks so dope! It really brings me back to the good old days of the internet, where the focus was purely on community and content without the clutter of ads. It’s so refreshing to see people building something that really feels like the internet used to.
#pixelfed
#mastodon

The Green Party of England and Wales now has more members than the Lib Dems
https://bright-green.org/2025/10/04/the-green-party-now-has-more-members-than-the-lib-dems/

The Green Party now has more members than the Lib Dems
There are now more than 83,500 Green Party members.

Anzahl der #Balkonkraftwerke je Bundesland pro 10000 Einwohner in #Deutschland mit Stand vom 05.09.2025.
👉 Zusatzlesestoff: Steckersolaranlage überwachen

#EU5 preloading.. I can feel the next year of my life slipping away already :)

Wie cool wäre das, wenn man bei "Standort" in einen Termin die Homepage von etwas eintragen könnte und es gäbe eine ".well-known/location" die geparst würde und in allen Apps funktionieren würde.

"The report said 83% of all countries, accounting for 90% of the world’s population, met the World Bank’s definition of high inequality. It added that countries with high inequality were seven times more likely to experience democratic decline than more equal countries."
Good lord, you don't say. Who would have thought it?
#Kleptocracy

A knowledgeable obituary of
Jack DeJohnette, versatile jazz drummer known for Miles Davis fusion recordings, died aged 83 | #Jazz | The Guardian
https://www.…

On synchronized coded systems
Wolfgang Krieger
https://arxiv.org/abs/2509.04122 https://arxiv.org/pdf/2509.04122 …

On synchronized coded systems
We introduce a class of codes with overlapping code words, that we call SPO-codes. The SPO-codes are related to the Markov codes that were introduced in: G. Keller, J. Combinatorial Theory 56, (1991),pp.\ 75--83. The process of generating a coded system from a code extends to SPO-codes. We describe a family of intrinsically ergodic synchronized SPO-coded systems that is closed under topological conjugacy. We construct synchronized subshifts with salient structural features by means of SPO-codes…

PV Image Sousou no Frieren Season 2 (Frieren: Beyond Journey's End Season 2)
Premieres January 16 2026 #フリーレン2期 #フリーレン #frieren

🇺🇦 #NowPlaying on KEXP's #MechanicalBreakdown
Decena Trágica:
🎵 Entre Sombras
#DecenaTrágica
https://open.spotify.com/track/2Pe9CgjeCJ2k73fml93hEL

084 Парк "Геєнна"
#comic #ТемнаНаука #переклади

@… Don't suppose you got a code for EU5? Very interested to find out if it runs okay in Linux!

Ich: "... schau mal bitte, wann dir das in den Kalender passt."
Er: "Also Samstag bin ich in $EntfernterOrt. Und Sonntag mit den Kindern bei $Veranstaltung."
Du sollst nicht gucken, wann es dir *nicht* passt, Alter, ey. Beantworte die Frage! 🤬

aktuelles #Wetter #Behrendorf #Wische 06-11-2025 11:25
Temperatur: (min: 2,5C) 10,8C (max: 10,9C)
Feuchte: 83%
Luftdruck: 1013,3 hPa (langsam fallend)
Windrichtung: 185 Grad
Wi…

Granskning: 83 procent av de dödade i Gaza var civila https://www.dn.se/sverige/granskning-83-procent-av-de-dodade-i-gaza-var-civila/

Antalet dödade i Gaza kan vara fler än man tidigare trott
Bland de palestinier som har dödats av den israeliska krigsmakten IDF utgör civila en mycket stor andel, uppger forskare från Uppsala…

Team dinners are a blast! We've been feasting on some seriously amazing food! Talk about a flavor explosion! I just indulged in the most delicious Asian catfish, and wow, I’m happily stuffed! I can't wait for the next feast!
#food
#dinner

[2025-09-03 Wed (UTC), 83 new articles found for cs.AI Artificial Intelligence]
toXiv_bot_toot

Dulin, S., Mitoulis, SA., Bredikhin, A. et al. 3 November 2025
"Rethinking infrastructure design from component failure to systemic resilience."
Nature Communications 16, 9681 (2025).
https://doi.org/10.1038/s41467-025-64683-6
pdf:

Die Folgen der perfekten Asymmetrie: mindestens 83% der Getöteten sind Zivilisten. In einem früheren Artikel hatte Yuval Abraham über das System Lavender berichtet, das #Palantir für die israelische Armee bereit stellt, unterstützt von der #KillCloud von

“ Meer dan de helft van ons land bestaat uit landbouwgrond (~54%), waarvan ruim een derde (~35%) bestaat uit grasland en voedergewassen die in hoofdzaak voor veevoer worden gebruikt. Het vlees dat daarvan wordt geproduceerd, verdwijnt voor 60-70% over de grens. En dit levert economisch minder dan één procent bij aan ons BBP. ”

𝐃𝐞 𝐤𝐨𝐞 𝐢𝐧 𝐝𝐞 𝐤𝐚𝐦𝐞𝐫 𝐰𝐨𝐫𝐝𝐭 𝐯𝐚𝐤𝐤𝐮𝐧𝐝𝐢𝐠 𝐠𝐞𝐧𝐞𝐠𝐞𝐞𝐫𝐝 We horen vaak dat Nederland vol is. Dat er geen ruimte meer is voor woningen, of dat migratie het probleem is. Het… | Norbert Schotte
𝐃𝐞 𝐤𝐨𝐞 𝐢𝐧 𝐝𝐞 𝐤𝐚𝐦𝐞𝐫 𝐰𝐨𝐫𝐝𝐭 𝐯𝐚𝐤𝐤𝐮𝐧𝐝𝐢𝐠 𝐠𝐞𝐧𝐞𝐠𝐞𝐞𝐫𝐝 We horen vaak dat Nederland vol is. Dat er geen ruimte meer is voor woningen, of dat migratie het probleem is. Het bekt waarschijnlijk lekker, doet het aardig voor electoraal gewin, maar wie verder kijkt dan bepaalde populistische politici, ziet een ander verhaal. Meer dan de helft van ons land bestaat uit landbouwgrond (~54%), waarvan rui…

Via @… I found this interesting constituency dashboard:
https://open-innovations.github.io/constituency-dashboards/
It's in…

芳川ヨシノ – Timeless Echoes
#pop

#まいにちカレー部

"Based on latest data that’s more than 360,000 innocent people murdered"
Israeli military’s own data says at least 83% of its Gaza victims were civilians – SKWAWKBOX
https://skwawkbox.org/2025/08/22/israeli-militarys-own-data-says-at-least-83-of-its-gaza-victims-were-civilians/

Replaced article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/1]:
- Learning to sample fibers for goodness-of-fit testing
Ivan Gvozdanovi\'c, Sonja Petrovi\'c

Я раньше и не подозревала, что Википедия выдаёт платину в промышленных масштабах
https://ru.wikipedia.org/wiki/Википедия:Форум/Вниманию_участников#Иллюстрация_статьи_«Скуф»

I recently ish played a text adventure mystery logic puzzle game where there was a big gathering at a house and people keep dying. There were gunshots off screen. I want to recommend it but can't recall it's name or where I played it ... Maybe in a web browser? Anyone?

@… duuhhhh 🥺🥺🥺

🇺🇦 #NowPlaying on #BBC6Music's #NewMusicFix
Турбо дизель:
🎵 Sweat It
#Турбодизель

Nach 25 Jahren endlich Riven (the sequel to Myst) geschafft.
Letztlich habe ich es natürlich erst wieder letztes Jahr aufgegriffen, als es neu raus kam, aber schon mega toll endlich einen Abschluss zu finden.

aktuelles #Wetter #Behrendorf #Wische 06-09-2025 20:30
Temperatur: (min: 9,3C) 16,0C (max: 23,5C)
Feuchte: 83%
Luftdruck: 1023,5 hPa (gleichbleibend)
Windrichtung: N/A Grad
…

🇺🇦 #NowPlaying on KEXP's #Expansions
Guedra Guedra كدرة كدرة:
🎵 Paradigm
#GuedraGuedraكدرةكدرة
https://guedraguedra.bandcamp.com/track/paradigm

Bernd das Brot ist in der aktuellen Folge von "Last Week Tonight" falls das jemanden interessiert. :-)

India's Ministry of Information and Broadcasting orders the takedown of 138 YouTube videos and 83 Instagram posts on Adani Enterprises, citing a court order (The News Minute)
https://www.thenewsminute.com/news/govt-or

Govt orders takedown of 138 YouTube videos, 83 Instagram posts on Adani citing court order
Follow TNM’s WhatsApp channel for news updates and story links.The Ministry of Information and Broadcasting (MIB) has issued takedown notices to sever

Hill returns to Dolphins' facility, speaks to team https://www.espn.com/nfl/story/_/id/46477683/tyreek-hill-returns-dolphins-facility-speaks-team

aktuelles #Wetter #Behrendorf #Wische 06-10-2025 15:50
Temperatur: (min: 10,3C) 15,5C (max: 15,8C)
Feuchte: 83%
Luftdruck: 1015,9 hPa (gleichbleibend)
Windrichtung: 284 Grad
Wi…

Arte klärt die wichtigen Fragen!
https://youtu.be/HgeDAJaPLs0

If you know, you know.
#думкиВголос #політика

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/1]:
- Replicable Clustering
Hossein Esfandiari, Amin Karbasi, Vahab Mirrokni, Grigoris Velegkas, Felix Zhou

つまり、日本での“スパイアンドスパイ”という名前は、ローカライズ担当者による命名と、日本のユーザーに分かりやすくする意図から生まれ、広告などで拡散され定着したことが理由です。
情報源
[1] 南国指令!!スパイvsスパイとは？ わかりやすく解説 https://www.weblio.jp/content/南国指令!!スパイvsスパイ
[2] 南国指令!!スパイvsスパイ

🇺🇦 #NowPlaying on #BBC6Music's #LaurenLaverne
Guedra Guedra كدرة كدرة:
🎵 Tamayyurt (feat. Foulane Bouhssine)
#GuedraGuedraكدرةكدرة
https://guedraguedra.bandcamp.com/track/tamayyurt

aktuelles #Wetter #Behrendorf #Wische 06-10-2025 13:45
Temperatur: (min: 10,3C) 15,0C (max: 15,4C)
Feuchte: 83%
Luftdruck: 1015,7 hPa (langsam steigend)
Windrichtung: 305 Grad
…

🇺🇦 #NowPlaying on KEXP's #AfternoonShow
シートベルツ:
🎵 Tank!
#シートベルツ

I do not understand how #2FA on my company laptop works. At random, far-apart times a nondescript dialogue pops up out of nowhere asking me to enter a number in the authenticator app.
Why not ask for the second factor when I log into Windows (the first time of the day)? At least when there is no (known) network?
It also feels like it would not be terribly hard to produce something like…

Week 3 NFL player props, odds, picks: Back Saquon Barkley Over 83.5 rushing yards with Sunday NFL prop picks

https://www.cbssports.com/nfl/news/week-3-

Contraction and entropy production in continuous-time Sinkhorn dynamics
Anand Srinivasan, Jean-Jacques Slotine
https://arxiv.org/abs/2510.12639 https://arx…

Contraction and entropy production in continuous-time Sinkhorn dynamics
Recently, the vanishing-step-size limit of the Sinkhorn algorithm at finite regularization parameter $\varepsilon$ was shown to be a mirror descent in the space of probability measures. We give $L^2$ contraction criteria in two time-dependent metrics induced by the mirror Hessian, which reduce to the coercivity of certain conditional expectation operators. We then give an exact identity for the entropy production rate of the Sinkhorn flow, which was previously known only to be nonpositive. Exam…

AWS ist down, aber SchildiChat geht, weil das matrix drunter self-host ist. (Tuwunel sei dank) ~~~

というわけで、カレーとマネチキ（鶏の唐揚げ）です。
#TAMAコミ #TAMAコミ11

I ditched all my smart gadgets at home and replaced them with USB-C rechargeable lights. I don’t have any assistants, and I replaced my Kindle with a pocketbook.
Removing some technology from my home and daily life feels refreshing.
#tech
#smarthome

Hier schieben sich die Ringeltauben wieder die Eicheln am Stück und im Dutzend in's Gesicht.

Adapting Noise to Data: Generative Flows from 1D Processes
Jannis Chemseddine, Gregor Kornhardt, Richard Duong, Gabriele Steidl
https://arxiv.org/abs/2510.12636 https://

Adapting Noise to Data: Generative Flows from 1D Processes
We introduce a general framework for constructing generative models using one-dimensional noising processes. Beyond diffusion processes, we outline examples that demonstrate the flexibility of our approach. Motivated by this, we propose a novel framework in which the 1D processes themselves are learnable, achieved by parameterizing the noise distribution through quantile functions that adapt to the data. Our construction integrates seamlessly with standard objectives, including Flow Matching an…

Week 3 NFL player props, picks, odds: Back Saquon Barkley Over 83.5 rushing yards with Sunday NFL prop picks

https://www.cbssports.com/nfl/news/week-3-

Stanley Cups sind bei den Grundschülern in MeckPom angekommen.

Last Sunday, while sipping my morning brew, an impulsive light bulb went off in my head. It said, "You will ride a bike in Denmark; you will ride to three nearby cities."
I am not a sporty guy whatsoever, but it sounds fun… on paper.
I am unprepared to ride 150 km in a few days, and I expect to die 500 times an hour.
#coffee

Universal Adaptive Environment Discovery
Madi Matymov, Ba-Hien Tran, Maurizio Filippone
https://arxiv.org/abs/2510.12547 https://arxiv.org/pdf/2510.12547…

Universal Adaptive Environment Discovery
An open problem in Machine Learning is how to avoid models to exploit spurious correlations in the data; a famous example is the background-label shortcut in the Waterbirds dataset. A common remedy is to train a model across multiple environments; in the Waterbirds dataset, this corresponds to training by randomizing the background. However, selecting the right environments is a challenging problem, given that these are rarely known a priori. We propose Universal Adaptive Environment Discovery …

Today in starling-el, the ability to set spending categories! Our first non-read operation. It's much nicer doing this in #emacs IMO!

Switching to @… continues to look like a good idea.
https://fosstodon.org/@organicmaps/115202657995358203

Geopolitics, Geoeconomics and Risk:A Machine Learning Approach
Alvaro Ortiz, Tomasa Rodrigo
https://arxiv.org/abs/2510.12416 https://arxiv.org/pdf/2510.124…

Geopolitics, Geoeconomics and Risk:A Machine Learning Approach
We introduce a novel high-frequency daily panel dataset of both markets and news-based indicators -- including Geopolitical Risk, Economic Policy Uncertainty, Trade Policy Uncertainty, and Political Sentiment -- for 42 countries across both emerging and developed markets. Using this dataset, we study how sentiment dynamics shape sovereign risk, measured by Credit Default Swap (CDS) spreads, and evaluate their forecasting value relative to traditional drivers such as global monetary policy and m…

Improved Central Limit Theorem and Bootstrap Approximations for Linear Stochastic Approximation
Bogdan Butyrin, Eric Moulines, Alexey Naumov, Sergey Samsonov, Qi-Man Shao, Zhuo-Song Zhang
https://arxiv.org/abs/2510.12375

Improved Central Limit Theorem and Bootstrap Approximations for Linear Stochastic Approximation
In this paper, we refine the Berry-Esseen bounds for the multivariate normal approximation of Polyak-Ruppert averaged iterates arising from the linear stochastic approximation (LSA) algorithm with decreasing step size. We consider the normal approximation by the Gaussian distribution with covariance matrix predicted by the Polyak-Juditsky central limit theorem and establish the rate up to order $n^{-1/3}$ in convex distance, where $n$ is the number of samples used in the algorithm. We also prov…

Dass sie einem im toom-Baumarkt an der Kasse "Viel Spaß damit!" wünschen kommt von Corporate, oder?
War aber trotzdem witzig zu hören. Hab nen WC-Sitz gekauft.

Learning Latent Energy-Based Models via Interacting Particle Langevin Dynamics
Joanna Marks, Tim Y. J. Wang, O. Deniz Akyildiz
https://arxiv.org/abs/2510.12311 https://

Learning Latent Energy-Based Models via Interacting Particle Langevin Dynamics
We develop interacting particle algorithms for learning latent variable models with energy-based priors. To do so, we leverage recent developments in particle-based methods for solving maximum marginal likelihood estimation (MMLE) problems. Specifically, we provide a continuous-time framework for learning latent energy-based models, by defining stochastic differential equations (SDEs) that provably solve the MMLE problem. We obtain a practical algorithm as a discretisation of these SDEs and pro…

Als ich die Überwachungskamera-Bilder von ihm sah, dachte ich: Ja nee, Schnucki - Basecap und Sonnenbrille werden deine Bekannten nicht davon abhalten, dich wieder zu erkennen.

Follow-the-Perturbed-Leader for Decoupled Bandits: Best-of-Both-Worlds and Practicality
Chaiwon Kim, Jongyeong Lee, Min-hwan Oh
https://arxiv.org/abs/2510.12152 https://

Follow-the-Perturbed-Leader for Decoupled Bandits: Best-of-Both-Worlds and Practicality
We study the decoupled multi-armed bandit (MAB) problem, where the learner selects one arm for exploration and one arm for exploitation in each round. The loss of the explored arm is observed but not counted, while the loss of the exploited arm is incurred without being observed. We propose a policy within the Follow-the-Perturbed-Leader (FTPL) framework using Pareto perturbations. Our policy achieves (near-)optimal regret regardless of the environment, i.e., Best-of-Both-Worlds (BOBW): constan…

Compressibility Measures Complexity: Minimum Description Length Meets Singular Learning Theory
Einar Urdshals, Edmund Lau, Jesse Hoogland, Stan van Wingerden, Daniel Murfet
https://arxiv.org/abs/2510.12077

Compressibility Measures Complexity: Minimum Description Length Meets Singular Learning Theory
We study neural network compressibility by using singular learning theory to extend the minimum description length (MDL) principle to singular models like neural networks. Through extensive experiments on the Pythia suite with quantization, factorization, and other compression techniques, we find that complexity estimates based on the local learning coefficient (LLC) are closely, and in some cases, linearly correlated with compressibility. Our results provide a path toward rigorously evaluating…

Statistical Guarantees for High-Dimensional Stochastic Gradient Descent
Jiaqi Li, Zhipeng Lou, Johannes Schmidt-Hieber, Wei Biao Wu
https://arxiv.org/abs/2510.12013 https://

Statistical Guarantees for High-Dimensional Stochastic Gradient Descent
Stochastic Gradient Descent (SGD) and its Ruppert-Polyak averaged variant (ASGD) lie at the heart of modern large-scale learning, yet their theoretical properties in high-dimensional settings are rarely understood. In this paper, we provide rigorous statistical guarantees for constant learning-rate SGD and ASGD in high-dimensional regimes. Our key innovation is to transfer powerful tools from high-dimensional time series to online learning. Specifically, by viewing SGD as a nonlinear autoregres…

Simplifying Optimal Transport through Schatten-$p$ Regularization
Tyler Maunu
https://arxiv.org/abs/2510.11910 https://arxiv.org/pdf/2510.11910

Simplifying Optimal Transport through Schatten-$p$ Regularization
We propose a new general framework for recovering low-rank structure in optimal transport using Schatten-$p$ norm regularization. Our approach extends existing methods that promote sparse and interpretable transport maps or plans, while providing a unified and principled family of convex programs that encourage low-dimensional structure. The convexity of our formulation enables direct theoretical analysis: we derive optimality conditions and prove recovery guarantees for low-rank couplings and …

High-Probability Bounds For Heterogeneous Local Differential Privacy
Maryam Aliakbarpour, Alireza Fallah, Swaha Roy, Ria Stevens
https://arxiv.org/abs/2510.11895 https://…

High-Probability Bounds For Heterogeneous Local Differential Privacy
We study statistical estimation under local differential privacy (LDP) when users may hold heterogeneous privacy levels and accuracy must be guaranteed with high probability. Departing from the common in-expectation analyses, and for one-dimensional and multi-dimensional mean estimation problems, we develop finite sample upper bounds in $\ell_2$-norm that hold with probability at least $1-β$. We complement these results with matching minimax lower bounds, establishing the optimality (up to con…

Active Subspaces in Infinite Dimension
Poorbita Kundu, Nathan Wycoff
https://arxiv.org/abs/2510.11871 https://arxiv.org/pdf/2510.11871

Active Subspaces in Infinite Dimension
Active subspace analysis uses the leading eigenspace of the gradient's second moment to conduct supervised dimension reduction. In this article, we extend this methodology to real-valued functionals on Hilbert space. We define an operator which coincides with the active subspace matrix when applied to a Euclidean space. We show that many of the desirable properties of Active Subspace analysis extend directly to the infinite dimensional setting. We also propose a Monte Carlo procedure and discus…

On Thompson Sampling and Bilateral Uncertainty in Additive Bayesian Optimization
Nathan Wycoff
https://arxiv.org/abs/2510.11792 https://arxiv.org/pdf/2510.…

On Thompson Sampling and Bilateral Uncertainty in Additive Bayesian Optimization
In Bayesian Optimization (BO), additive assumptions can mitigate the twin difficulties of modeling and searching a complex function in high dimension. However, common acquisition functions, like the Additive Lower Confidence Bound, ignore pairwise covariances between dimensions, which we'll call \textit{bilateral uncertainty} (BU), imposing a second layer of approximations. While theoretical results indicate that asymptotically not much is lost in doing so, little is known about the practical e…

Dimension-Free Minimax Rates for Learning Pairwise Interactions in Attention-Style Models
Shai Zucker, Xiong Wang, Fei Lu, Inbar Seroussi
https://arxiv.org/abs/2510.11789 https:…

Dimension-Free Minimax Rates for Learning Pairwise Interactions in Attention-Style Models
We study the convergence rate of learning pairwise interactions in single-layer attention-style models, where tokens interact through a weight matrix and a non-linear activation function. We prove that the minimax rate is $M^{-\frac{2β}{2β+1}}$ with $M$ being the sample size, depending only on the smoothness $β$ of the activation, and crucially independent of token count, ambient dimension, or rank of the weight matrix. These results highlight a fundamental dimension-free statistical efficie…

[2025-10-15 Wed (UTC), 15 new articles found for stat.ML Machine Learning]
toXiv_bot_toot

Dendrograms of Mixing Measures for Softmax-Gated Gaussian Mixture of Experts: Consistency without Model Sweeps
Do Tien Hai, Trung Nguyen Mai, TrungTin Nguyen, Nhat Ho, Binh T. Nguyen, Christopher Drovandi
https://arxiv.org/abs/2510.12744

Dendrograms of Mixing Measures for Softmax-Gated Gaussian Mixture of Experts: Consistency without Model Sweeps
We develop a unified statistical framework for softmax-gated Gaussian mixture of experts (SGMoE) that addresses three long-standing obstacles in parameter estimation and model selection: (i) non-identifiability of gating parameters up to common translations, (ii) intrinsic gate-expert interactions that induce coupled differential relations in the likelihood, and (iii) the tight numerator-denominator coupling in the softmax-induced conditional density. Our approach introduces Voronoi-type loss f…

Replaced article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[2/2]:
- Learning Shared Representations from Unpaired Data
Amitai Yacobi, Nir Ben-Ari, Ronen Talmon, Uri Shaham

Replaced article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/2]:
- Sparse Robust Classification via the Kernel Mean
Brendan van Rooyen, Aditya Krishna Menon, Robert C. Williamson

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[3/3]:
- An Eulerian Perspective on Straight-Line Sampling
Panos Tsimpos, Youssef Marzouk

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[2/3]:
- What Makes Looped Transformers Perform Better Than Non-Recursive Ones (Provably)
Zixuan Gong, Jiaye Teng, Yong Liu

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/3]:
- Performance of Machine Learning Methods for Gravity Inversion: Successes and Challenges
Vahid Negahdari, Shirin Samadi Bahrami, Seyed Reza Moghadasi, Mohammad Reza Razvan

Efficient Group Lasso Regularized Rank Regression with Data-Driven Parameter Determination
Meixia Lin, Meijiao Shi, Yunhai Xiao, Qian Zhang
https://arxiv.org/abs/2510.11546 http…

Efficient Group Lasso Regularized Rank Regression with Data-Driven Parameter Determination
High-dimensional regression often suffers from heavy-tailed noise and outliers, which can severely undermine the reliability of least-squares based methods. To improve robustness, we adopt a non-smooth Wilcoxon score based rank objective and incorporate structured group sparsity regularization, a natural generalization of the lasso, yielding a group lasso regularized rank regression method. By extending the tuning-free parameter selection scheme originally developed for the lasso, we introduce …

PAC-Bayesian Bounds on Constrained f-Entropic Risk Measures
Hind Atbir, Farah Cherfaoui, Guillaume Metzler, Emilie Morvant, Paul Viallard
https://arxiv.org/abs/2510.11169 https:…

PAC-Bayesian Bounds on Constrained f-Entropic Risk Measures
PAC generalization bounds on the risk, when expressed in terms of the expected loss, are often insufficient to capture imbalances between subgroups in the data. To overcome this limitation, we introduce a new family of risk measures, called constrained f-entropic risk measures, which enable finer control over distributional shifts and subgroup imbalances via f-divergences, and include the Conditional Value at Risk (CVaR), a well-known risk measure. We derive both classical and disintegrated PAC…

torchsom: The Reference PyTorch Library for Self-Organizing Maps
Louis Berthier, Ahmed Shokry, Maxime Moreaud, Guillaume Ramelet, Eric Moulines
https://arxiv.org/abs/2510.11147 …

torchsom: The Reference PyTorch Library for Self-Organizing Maps
This paper introduces torchsom, an open-source Python library that provides a reference implementation of the Self-Organizing Map (SOM) in PyTorch. This package offers three main features: (i) dimensionality reduction, (ii) clustering, and (iii) friendly data visualization. It relies on a PyTorch backend, enabling (i) fast and efficient training of SOMs through GPU acceleration, and (ii) easy and scalable integrations with PyTorch ecosystem. Moreover, torchsom follows the scikit-learn API for e…

Adversarial Robustness in One-Stage Learning-to-Defer
Yannis Montreuil, Letian Yu, Axel Carlier, Lai Xing Ng, Wei Tsang Ooi
https://arxiv.org/abs/2510.10988 https://

Adversarial Robustness in One-Stage Learning-to-Defer
Learning-to-Defer (L2D) enables hybrid decision-making by routing inputs either to a predictor or to external experts. While promising, L2D is highly vulnerable to adversarial perturbations, which can not only flip predictions but also manipulate deferral decisions. Prior robustness analyses focus solely on two-stage settings, leaving open the end-to-end (one-stage) case where predictor and allocation are trained jointly. We introduce the first framework for adversarial robustness in one-stage …

Replaced article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/2]:
- Multiparameter regularization and aggregation in the context of polynomial functional regression
Gizewski, Holzleitner, Mayer-Suess, Pereverzyev, Pereverzyev

Crosslisted article(s) found for stat.ML. https://arxiv.org/list/stat.ML/new
[1/1]:
- Energy-Driven Steering: Reducing False Refusals in Large Language Models
Jiang, Ou, Liu, Pang, Wan, Duan, Dong, Chang, Wang, Wu, Li

Conditional Flow Matching for Bayesian Posterior Inference
So Won Jeong, Percy S. Zhai, Veronika Ro\v{c}ov\'a
https://arxiv.org/abs/2510.09534 https://…

Conditional Flow Matching for Bayesian Posterior Inference
We propose a generative multivariate posterior sampler via flow matching. It offers a simple training objective, and does not require access to likelihood evaluation. The method learns a dynamic, block-triangular velocity field in the joint space of data and parameters, which results in a deterministic transport map from a source distribution to the desired posterior. The inverse map, named vector rank, is accessible by reversibly integrating the velocity over time. It is advantageous to levera…

Interpretable Generative and Discriminative Learning for Multimodal and Incomplete Clinical Data
Albert Belenguer-Llorens, Carlos Sevilla-Salcedo, Janaina Mourao-Miranda, Vanessa G\'omez-Verdejo
https://arxiv.org/abs/2510.09513

Interpretable Generative and Discriminative Learning for Multimodal and Incomplete Clinical Data
Real-world clinical problems are often characterized by multimodal data, usually associated with incomplete views and limited sample sizes in their cohorts, posing significant limitations for machine learning algorithms. In this work, we propose a Bayesian approach designed to efficiently handle these challenges while providing interpretable solutions. Our approach integrates (1) a generative formulation to capture cross-view relationships with a semi-supervised strategy, and (2) a discriminati…

Efficient Autoregressive Inference for Transformer Probabilistic Models
Conor Hassan, Nasrulloh Loka, Cen-You Li, Daolang Huang, Paul E. Chang, Yang Yang, Francesco Silvestrin, Samuel Kaski, Luigi Acerbi
https://arxiv.org/abs/2510.09477

Efficient Autoregressive Inference for Transformer Probabilistic Models
Transformer-based models for amortized probabilistic inference, such as neural processes, prior-fitted networks, and tabular foundation models, excel at single-pass marginal prediction. However, many real-world applications, from signal interpolation to multi-column tabular predictions, require coherent joint distributions that capture dependencies between predictions. While purely autoregressive architectures efficiently generate such distributions, they sacrifice the flexible set-conditioning…

A unified Bayesian framework for adversarial robustness
Pablo G. Arce, Roi Naveiro, David R\'ios Insua
https://arxiv.org/abs/2510.09288 https://arxiv.o…

A unified Bayesian framework for adversarial robustness
The vulnerability of machine learning models to adversarial attacks remains a critical security challenge. Traditional defenses, such as adversarial training, typically robustify models by minimizing a worst-case loss. However, these deterministic approaches do not account for uncertainty in the adversary's attack. While stochastic defenses placing a probability distribution on the adversary exist, they often lack statistical rigor and fail to make explicit their underlying assumptions. To reso…

Distributionally robust approximation property of neural networks
Mihriban Ceylan, David J. Pr\"omel
https://arxiv.org/abs/2510.09177 https://arxiv.or…

Distributionally robust approximation property of neural networks
The universal approximation property uniformly with respect to weakly compact families of measures is established for several classes of neural networks. To that end, we prove that these neural networks are dense in Orlicz spaces, thereby extending classical universal approximation theorems even beyond the traditional $L^p$-setting. The covered classes of neural networks include widely used architectures like feedforward neural networks with non-polynomial activation functions, deep narrow netw…

Mirror Flow Matching with Heavy-Tailed Priors for Generative Modeling on Convex Domains
Yunrui Guan, Krishnakumar Balasubramanian, Shiqian Ma
https://arxiv.org/abs/2510.08929 ht…

Mirror Flow Matching with Heavy-Tailed Priors for Generative Modeling on Convex Domains
We study generative modeling on convex domains using flow matching and mirror maps, and identify two fundamental challenges. First, standard log-barrier mirror maps induce heavy-tailed dual distributions, leading to ill-posed dynamics. Second, coupling with Gaussian priors performs poorly when matching heavy-tailed targets. To address these issues, we propose Mirror Flow Matching based on a \emph{regularized mirror map} that controls dual tail behavior and guarantees finite moments, together wi…