Tootfinder

Google plans to integrate Nano Banana and Veo support into Gemini on Google TV, letting users generate AI videos and modify family photos, first on TCL TVs (Terrence O'Brien/The Verge)
https://www.theverge.com/tech/854112/gemini-google-tv-nano-banana-veo-ces

Like global search and replace but don’t like surprises?
Check out serpl – a handy little command-line app that gives you a visual preview of the changes you are about to make. You can even go in and remove the replacements you don’t want from the source previews. The regex support appears to be basic, however (I couldn’t get a negative lookbehind to work).

Because I constantly hear myths about the good old compact cassette here's a longer post dispelling them:
1. They can sound as good as CDs
2. They don't wear out
3. You can't use a pencil to wind them
4. You can go to specific tracks automatically
5. You don't need to carry around extra batteries
I will elaborate below:
1. Sound Quality
Many higher-end decks can record cassettes on metal tape with various Dolby noise reduction settings; especially the combination of metal tape and Dolby S will make tapes that are pretty much indistinguishable from listening to a CD.
Even normal or chrome tape with Dolby B (around since the 1970s) will give great results; likely indistinguishable from a CD when played in a car or while out and about with a personal player.
Some extremely high-end tape decks produce better than CD results in some regards (for example some Nakamichi models go to 26KHz with frequency response, while CD are inherently limited to top out at 22KHz).
It's true that the dynamic range of CDs is much better than either vinyl records or tapes. However, unless you're super into classical music there's likely not much music for which this truly matters, as 99% is mastered to use much less dynamic range than provided by any audio media format. (If you're super into classical music you probably want SACD or other high-res lossless sources anyway, not CDs.)
2. Yes, it will wear out mechanically but you will wear out mechanically before it does. Please watch VWestlife's video: https://www.youtube.com/watch?v=_dgJ4hRHBiw
3. European and American pencils are too thin to engage the cassette reel cogs. (You'd need to get a Japanese pencil. People mostly used BIC pens for this purpose which have the right thickness.)
4. Most (nice) decks and personal players from the early-to-mid nineties onwards have track skip features (e.g. Sony has AMS, Automatic Music Sensor), which allow precise winding to a specific track.
Some decks even did this in the early 80s!
5. My late-90s Walkman has seventy-eight (78) hours of playback on one (1) single AA battery.
Anyway, the main reason why I like them is they're fun to use and recording them is very deliberate instead of algorithms selecting music for me. :)

Brought home some #espresso beans. Pulled a shot on a Bambino Plus with default settings, then made a cappuccino on the same machine. Tomorrow I'll try with the Flair 58.
Thanks to @… for the lovely mug (it’s not blurry IRL).

Still getting used to all this 'vinyl' stuff; think I've managed to capture one side; it took quite a few goes on the last track to stop it getting stuck; now I got it through on about the 5th take of that track, but hmm - was it me tweaking the weight settings on the turntable, giving it another clean, or was it just getting happier as it actually had been played and fixing up small problems? Heck I dunno.
Got to say it sounds great on my dads Hifi.

Um, I was so fed up with multiplayer shooters, I kind of missed The Finals. It's pretty okay...ish I guess? Also, it just works on Linux, even after I threw FSR 4.1.0 at it (FSR 4 Quality, ~100 FPS, Max settings adaptive RT on the RX 7800 XT).

I saw a video where a guy built a Power BI app to filter only very specific senders for a client who got a lot of spam email. I commented, "there's a checkbox in OWA settings to only deliver emails from your safe senders list." Could've done it in like 2 clicks.

How to Block the ‘Upgrade to Tahoe’ Alerts and System Settings Indicator
https://daringfireball.net/linked/2026/02/27/how-to-block-the-upgrade-to-tahoe-alerts-and-system-settings-indicator
It is really annoy…

#Sunstone #browser grew a couple of features over the past two days.
- History
This is in the same `places` database as bookmarks. For the time being, there is no interface for browsing history or deleting entries. I've added two settings, max_entries and max_days. Every fiftee…

Does anybody know, by any rare chance, what #Firefox settings might cause CORS errors? Since last week I'm unable to access, for example, a local #Jellyfin instance with Firefox due to this problem, as it causes a lot of CORS errors (same origin policy).
I have already tried changing &quo…

#Firefox führt ab Version 148 eine neue Einstellung ein, mit der Nutzer alle aktuellen und zukünftigen generativen #AI-Funktionen im Browser blockieren können.
Alternativ lassen sich einzelne Features wie #Übersetzungen

@… thanks!
My gut feeling, which might be wrong (QEMU is very new to me): QEMU could be tricky – if the settings that are typical on, for example, Windows, macOS and Linux hosts differ substantially from settings that are typical for a FreeBSD 15 or 16 host.
I mean, don't let QEMU be a distraction.
In particular: I don…

🔊 #NowPlaying on #BBCRadio3:
#ClassicalLive
- The Armed Man - A Mass for Peace from Germany
Al Ryan with concert recordings of distinctive settings of the church mass, today with music by Karl Jenkins, plus music from the recent Belfast International Chamber Music Festival.
Relisten now 👇
https://www.bbc.co.uk/programmes/m002rtgk

Instead of an "AI kill switch" like they promised, #Mozilla will provide "AI controls" buried deep in settings a #Firefox update which isn't out yet, until then the AI features which are already released and on by default can only be disabled by digging through about:co…

uhh, were users *asking* for this? https://support.mozilla.org/en-US/kb/remote-improvements

On the Complexity of Maximal/Closed Frequent Tree Mining for Bounded Height Trees
Kenta Komoto, Kazuhiro Kurita, Hirotaka Ono
https://arxiv.org/abs/2602.03436 https://arxiv.org/pdf/2602.03436 https://arxiv.org/html/2602.03436
arXiv:2602.03436v1 Announce Type: new
Abstract: In this paper, we address the problem of enumerating all frequent maximal/closed trees. This is a classical and central problem in data mining. Although many practical algorithms have been developed for this problem, its complexity under ``realistic assumptions'' on tree height has not been clarified. More specifically, while it was known that the mining problem becomes hard when the tree height is at least 60, the complexity for cases where the tree height is smaller has not yet been clarified. We resolve this gap by establishing results for these tree mining problems under several settings, including ordered and unordered trees, as well as maximal and closed variants.
toXiv_bot_toot

wild kubuntu seem to have the nextcloud desktop app in the "store" regular ubuntu did not.
I had to run:
sudo apt install nextcloud-desktop
the nextcloud website has you download some appImage file that I have no idea how to use.
also nextcloud on kubuntu was slow as ass. settings pane took 30 seconds to open.
#ubuntu

Wow #MicrosoftTeams web version somehow manages to completely bypass my audio settings. It'll just blare its sound out the display speakers, impossible to use my headphones. Great peace of software. /s

💥 Tiny sensor could transform head injury detection
#sensors

RE: https://hachyderm.io/@fantinel/115991351707305503
Well-written post, especially the “About Phanpy” part 😉
Very cool to see this.

As a "safety measure," Discord will require a face scan or ID starting next month.
https://discord.com/press-releases/discord-launches-teen-by-default-settings-globally

WhatsApp is spyware because its encryption is very questionable! The metadata being transferred openly reveals this very fell out over one. This is just marketing that their users should follow thoughtlessly.
»WhatsApp Rolls Out #Lockdown-Style Security Mode to Protect Targeted Users From #Spyware<…

RE: https://mastodon.social/@firefoxwebdevs/116002119945073671
I am vaguely annoyed these are called ‘enhancements’.
Animated GIF alt:
App menu − Settings − AI Controls − “Block AI Enhancements” (toggle).
Then scroll down and find …

I mean what?
What?
I don't have any idea how you'd embed 7 tablet screens into a semi-circular pine table, let alone also have some kind of telescopic central column?
Even if I knew how I don't have the money for seven individual in-console embedded tabled place-settings.
This is madness. I should be asleep.

🚀 Enable via new Plugins section at https://openrouter.ai/settings/plugins - opt-in toggle activates automatic healing layer for all structured output requests
💡 Key insight: A 2% defect rate dropping to 1% means cutting defects, bugs, and support tickets in half - reliability at the margins is wh…

The built-in database backup and restore feature in Kitten¹ (that actually works and is in the Kitten Settings section of every Kitten app) just saved my ass (again) :)
Thank you, past me ;)
¹ https://kitten.small-web.org

Had cause to visit, thought I'd share just in case
#FCPX #DaVinciResolve #Premiere #VideoEditing

We've been playing around with stream settings for the past week. Here are the best settings we've tested for streaming #LOTRO and other games from #Linux. They should work well on Windows as well. Remember, if you're on Nvidia, replace the "VAAPI" with "HVENC".

Source: the DOD appears to have accepted OpenAI's safety red lines, which were similar to Anthropic's, to deploy OpenAI's tech in classified settings (Maria Curi/Axios)
https://www.axios.com/2026/02/27/pentagon-openai-safety-red-lines-anthropic

#Discord Launches Teen-by-Default Settings Globally
https://discord.com/press-releases/discord-launches-teen-by-default-settings-globally

I donated to Partners for Health
https://www.pih.org/our-mission
"By establishing long-term relationships with sister organizations based in settings of poverty, Partners In Health strives to achieve two overarching goals: to bring the benefits of modern medical science to those most in need of…

PRs… not welcome? 😅
https://github.blog/changelog/2026-02-13-new-repository-settings-for-configuring-pull-request-access/

If I may, folks, if you’re trying to figure out remote control settings on your folks’ TVs, consider uploading a pic to your fave LLM to ask. Just worked like a charm for me.

WhatsApp launches Strict Account Settings, a high-security mode to protect high-risk users like journalists and public figures from sophisticated cyberattacks (Raphael Satter/Reuters)
https://www.reuters.com/business/media-tel

huh, based

[OT] Latest Firefox browser: block generative AI features with Firefox AI controls #AI

Interestingly a simple kWh/HDD regression for our full electricity imports fits reasonably well R^2>0.75) and gives a total year prediction close to that of our electricity supplier #Ecotricity partly because almost all imports are in winter for space heat...

#PhanpySocial changelog ✨
⌨️ More sequential hotkeys, g>s (Settings), g>p (Profile), g>b (Bookmarks)
💈 Redesigned poll
🛬 Redesigned landing page
🐛 Bug fixes
🔗 https://phanpy.social/…

Are you like me and avoiding MacOS 26 Tahoe and iOS 26 like the plague?
Opt-in to the beta programs for each of the previous version in Settings > General. Your Mac and iPhone/iPad will stop asking you to upgrade to the latest versions and you'll automatically receive security updates for Sequoia and iOS 18, respectively. You're welcome.
#apple

Did another sanity check on the 2nd PC and installed The Alters which is an Unreal 5 game. My internet bill gonna be higher this month cause I apparently passed the cap limit (thanks for failing to send me the warning email Comcast) but whatever. Booted it, checked the settings (Epic, Native AA so the 1080p the display runs at) and loaded in.. to a buttery smooth 60fps.
So it's definitely something going on with The Last Caretaker. I have no idea what, tho. For some reason in game …

Series C, Episode 12 - Death-Watch
TARRANT: She'll have had some help, of course, but not from the High Council of Vandor itself, or not deliberately.
AVON: There's an arms manufacturing cartel that looks highly promising as a candidate for co-conspirator.
https://blake.torpidity.net/m/312/596

Benchmarking of algorithms for set partitions
Arnav Khinvasara, Alexander Pikovski
https://arxiv.org/abs/2602.01350 https://arxiv.org/pdf/2602.01350 https://arxiv.org/html/2602.01350
arXiv:2602.01350v1 Announce Type: new
Abstract: Set partitions are arrangements of distinct objects into groups. The problem of listing all set partitions arises in a variety of settings, in particular in combinatorial optimization tasks. After a brief review, we give practical approximate formulas for determining the number of set partitions, both for small and large set sizes. Several algorithms for enumerating all set partitions are reviewed, and benchmarking tests were conducted. The algorithm of Djokic et al. is recommended for practical use.
toXiv_bot_toot

Customize what happens when you start R: #rstats #environment

Does someone here know how to restore #signal messages from the cloud #backup?
It seems to be the only way to create a backup on #iOS. Unfortunately, the option from the documentation does not exist on the new

High-Dimensional Robust Mean Estimation with Untrusted Batches
Maryam Aliakbarpour, Vladimir Braverman, Yuhan Liu, Junze Yin
https://arxiv.org/abs/2602.20698 https://arxiv.org/pdf/2602.20698 https://arxiv.org/html/2602.20698
arXiv:2602.20698v1 Announce Type: new
Abstract: We study high-dimensional mean estimation in a collaborative setting where data is contributed by $N$ users in batches of size $n$. In this environment, a learner seeks to recover the mean $\mu$ of a true distribution $P$ from a collection of sources that are both statistically heterogeneous and potentially malicious. We formalize this challenge through a double corruption landscape: an $\varepsilon$-fraction of users are entirely adversarial, while the remaining ``good'' users provide data from distributions that are related to $P$, but deviate by a proximity parameter $\alpha$.
Unlike existing work on the untrusted batch model, which typically measures this deviation via total variation distance in discrete settings, we address the continuous, high-dimensional regime under two natural variants for deviation: (1) good batches are drawn from distributions with a mean-shift of $\sqrt{\alpha}$, or (2) an $\alpha$-fraction of samples within each good batch are adversarially corrupted. In particular, the second model presents significant new challenges: in high dimensions, unlike discrete settings, even a small fraction of sample-level corruption can shift empirical means and covariances arbitrarily.
We provide two Sum-of-Squares (SoS) based algorithms to navigate this tiered corruption. Our algorithms achieve the minimax-optimal error rate $O(\sqrt{\varepsilon/n} \sqrt{d/nN} \sqrt{\alpha})$, demonstrating that while heterogeneity $\alpha$ represents an inherent statistical difficulty, the influence of adversarial users is suppressed by a factor of $1/\sqrt{n}$ due to the internal averaging afforded by the batch structure.
toXiv_bot_toot

I have a tablet whose main function is to play Polytopia in the metro. No Sim and most of the time no wi-fi either.
So I turned out all notifications
But now I get notification to tell me I can enable notification in settings...
I'm sick of this.

*click* Until Firefox devs screw something up that's absolutely no AI functionality for me, thanks.
Oh, it's in its own new section of the settings, "AI Controls".
#firefox #mozilla #noai

Got some time to spare.
One install, one git clone, two file copies and one rebuild and that whole system, with all my applications. tweaks and settings, was rebuilt in exactly 31 minutes. The power of reproducibilty.
I'll grab a beer and head out to that bench.
#nixos #cosmic…

If you manage your code on Github, you might expect that you get an email notification if someone opens an issue for your code. But it appears that's no longer true for new repos.
The "Watch" settings are now, by default, "Participating and Mentions". That means unless someone tags your Github username in the issue report, you won't know about it. Yeah, even if it's your own repo in your own namespace. No, I don't know what they were thinking at Githu…

Edit: thanks to @…, figured out the Musicolet button(s) to turn off auto-repeat on full albums. It's because they call them queues.
#Musicolet #Android

The official Epson Print Layout application is such an annoying piece of bad software, bug ridden and slowly turning out to be quite costly (when working with expensive fine art inks & paper, costing 2.50 EUR per page)...
Any recommendations for alternatives (which also support the P900 advanced ink settings?)
Just one example: I go through the effort to create a nice custom page layout in a certain orientation (e.g. A3 landscape). The app also shows the preview correctly, bu…

Sources: multiple federal agencies raised concerns about Grok's safety and reliability in recent months, before DOD approved Grok for use in classified settings (Wall Street Journal)
https://www.wsj.com/politics/national-security/elon…

By which, Linked means "you're going to have to click 6 times to disable bullshit notifications about games", of course.

#today I have discussed a possible research collaboration following my supervision yesterday, I am considering demand shaping from my heat pump with 'quiet' modes*, and I am ploughing through #NewScientist articles.
*

@… Tak - weszły - musiałem przerzucić urządzenie z .com do .it (znalazłem to na jakimś tickecie).

Czyli nie tylko "Country/Region Settings", ale też zrobić na ekrenie "Manage Your Content and Devices" transfer.

If ever a dev earned their keep, it's @…

Evaluating #AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?
https://arxiv.org/abs/2602.11988

🐀 Rats demonstrate ability to replay episodic memories in complex experimental settings
#psychology

@… @… Regarding the discussion on using MDM profiles to disable the Red Badge of Tahoe in System Settings, there is another workaround using Prefs Editor, which I found via Reddit (

Another block of a web site. I have autoplay video and sound off. Started reading the article and they started playing... I don't care how good you are if you ignore my settings then I permanently ignore you.

@… there is an option in the accessibility to render it less awfull :
"Reduce Transparency" in Settings > Accessibility > Display & Text Size
(I’m Still on ios18 but my girlfriend had her ipad on auto-update)

I have a major new feature for Left Wordle. I am testing the ability to sync settings and game history between devices. This feature is very much beta at this time and everything is subject to change.
My goal is to get as many people testing this as possible to identify any bugs that might cause problems once released. If you have time, please give it a try. Instructions here:

Reminder: You can let Mozilla know what you think about AI and other stupid shit.
https://connect.mozilla.org/t5/discussions/building-ai-the-firefox-way-shaping-what-s-next-together/td-p/109922
(Edited to fix li…

Let's expose more hidden links in #Jellyfin for #Roku!!!!!!!!
Moving the search and settings links out from under the user menu and putting them directly on the home screen.
#UX

JUST-DUB-IT: Video Dubbing via Joint Audio-Visual Diffusion
Anthony Chen, Naomi Ken Korem, Tavi Halperin, Matan Ben Yosef, Urska Jelercic, Ofir Bibi, Or Patashnik, Daniel Cohen-Or
https://arxiv.org/abs/2601.22143 https://arxiv.org/pdf/2601.22143 https://arxiv.org/html/2601.22143
arXiv:2601.22143v1 Announce Type: new
Abstract: Audio-Visual Foundation Models, which are pretrained to jointly generate sound and visual content, have recently shown an unprecedented ability to model multi-modal generation and editing, opening new opportunities for downstream tasks. Among these tasks, video dubbing could greatly benefit from such priors, yet most existing solutions still rely on complex, task-specific pipelines that struggle in real-world settings. In this work, we introduce a single-model approach that adapts a foundational audio-video diffusion model for video-to-video dubbing via a lightweight LoRA. The LoRA enables the model to condition on an input audio-video while jointly generating translated audio and synchronized facial motion. To train this LoRA, we leverage the generative model itself to synthesize paired multilingual videos of the same speaker. Specifically, we generate multilingual videos with language switches within a single clip, and then inpaint the face and audio in each half to match the language of the other half. By leveraging the rich generative prior of the audio-visual model, our approach preserves speaker identity and lip synchronization while remaining robust to complex motion and real-world dynamics. We demonstrate that our approach produces high-quality dubbed videos with improved visual fidelity, lip synchronization, and robustness compared to existing dubbing pipelines.
toXiv_bot_toot

Maybe someone can point me the right direction. I want to provision some default settings for LibreOffice on Linux, e.g. set the Ribbons-like interface. What's the proper way?
#libreoffice

I get a steady stream in my replies of meant-to-be-helpful advice of the form “You should use [complex F/OSS tech] to [do complex thing] that [solves a problem you don’t have but renders the process unusable for the people doing it]”
…when the real tech problems we’re facing on the ground here, both with Signal and other tools, are 99% just ground-level UX stuff:
- basic, basic usability glitches
- workflow friction
- problematic default settings
- reliability

the default drop down for "screen blank" is so weird. Also not sure why "screen blank" and not "screen lock".
here is how you set a custom time: https://askubuntu.com/questions/1042641/how-to-set-custom-lock-screen-time…

@… @…
SDDM, GNOME, X.Org.
Initially: a black screen on the display to the left.
For the screen to appear on the primary display: it was necessary to use settings on the secondary display, to temporarily make the sec…

Just the Browser helps you remove AI features, telemetry data reporting, sponsored content, product integrations, and other annoyances from desktop web browsers. The goal is to give you "just the browser" and nothing else, using hidden settings in web browsers intended for companies and other organizations.
https://justthebrowser.com/

Nvidia launches a native GeForce NOW app for Linux, starting with Debian-based Ubuntu, after releasing a Steam Deck app in 2025; in a hands-on, it works well (Ben Wilson/Windows Central)
https://www.windowscentral.com/gaming/pc-gaming/nvidia-geforce…

I was experiencing unusually high VRAM usage on the RX 7800 XT (GPU choking and cr@pping itself) in Ratchet & Clank: Rift Apart (and in a couple of other games). So I decided to use the AMD Cleanup Utility and reinstall the current 25.12.1 drivers. When I started the game again, the ray tracing settings in it were grayed out, and there's NO fixing it, no matter what I tried 😆 And the VRAM issue wasn't fixed 😄
(Grayed out RT settings aslo seem to be a problem on the 9000 s…

When even #Firefox ends up in my Mastodon filters. What a timeline. #FuckAI

Crossmodal interaction of flashes and beeps across time and number follows Bayesian causal inference #SiFI

Hey @… I just noticed that the new favicon is rather pixelated on higher DPI settings and now I can't unsee it. Any chance you could jump on the SVG favicon bandwagon?

Just finished "I Shall Never Fall in Love" by Hari Conner. An excellent graphic novel about some queer friends (and more-than-friends) in Georgian England.
I'm not usually a fan of such settings as the imperial harms abroad on which those societies are predicated make it harder for me to sympathize with the characters, but this book actually addresses that (if imperfectly) and there's enough nuance for me not to hate it, so I really enjoyed the cute romance.
I just wish there were more historical narratives written from lower-class perspectives that didn't make their characters seem unintelligent.
#AmReading #ReadingNow

Using MS365 for email is accepting the combined unknown unknowns of both MS operating policy choices and their other customers’ security.
And the KNOWN risk that it makes phishing much easier to execute in ways which are harder to catch.
#Sysadminnery @…

The new Raspberry Pi Connect feature works really, really well. After setting one up for work, it has me rethinking what I want to do with my various pieces of pi-ware at home. The latest Raspberry Pi Imager is a bit of a joke, though. Two different installs for two different machines where it completely ignored my custom settings during install, so I had to set up everything twice.

https://justthebrowser.com/
"Just the Browser helps you remove AI features, telemetry data reporting, sponsored content, product integrations, and other annoyances from desktop web browsers. The goal is to give you just the browser and nothing else, using hidden settings in web browsers intended for com…

A reminder to the peeps updating to iOS 26 this week:
Settings ➼ Display & Brightness ➼ Liquid Glass ➼ Tinted
Really takes it from almost unusable to just bad.

I am so, so tired of having to vet every one of my choices around a technology. Which platform? Which settings? What's a privacy risk? What has downstream negative social consequences? How do I compensate for the functionality I lose when I make a conscious choice to limit or refuse?
1/n
#rant

Mozilla think that adding some controls to their settings to allow users to opt out of the AI features is good enough.
No sign of a separate build without them there at all, but at least a switch to disable them, in theory.
They say:
"Once configured, the AI preferences will persist across browser updates"
Though they do have a history of turning on telemetry over and over again even when you keep turning it off. Or inventing new telemetery for you to turn off each update.
One switch will turn off:
"AI-enhanced tab grouping with suggested names,"
Which sounds horrifying, but then I don't really like tab groups anyway.
They seem to be suggesting that this is all just for desktop though, and so the Android build will presumably continue to not even have a working about:config at all.
https://blog.mozilla.org/en/firefox/ai-controls/
https://www.bleepingcomputer.com/news/software/mozilla-will-let-you-turn-off-all-firefox-ai-features/

These are the same picture.

Interesting 👀 https://social.zarchbox.fr/notes/agclhhyre66bnqz3
Extension from @… :

WhatsApp launches Strict Account Settings, a high-security mode to protect high-risk users like journalists and public figures from sophisticated cyberattacks (Raphael Satter/Reuters)
https://www.reuters.com/business/media-tel

'k3b' is a CD/DVD burner with audio ripper - now, I'd have to say I've not actually burnt a disc in many years, but I do still use the CD audio ripping. It's got some nice settings for retrying slightly tired discs.
https://apps.kde.org/en-gb/k3b/

On my 3rd attempt at activating my mother in law's new iPad. No, we don't want Touch ID; no we don't want a passcode. Process freezes here each time. Been at it for 45mins.
The idea of literally anyone else in the family pulling it off is laughable, never mind the elderly.
#iOS #ios26

It seems very appropriate to set the regional settings for English from United States to Kingdom of Denmark by now 🤣

Report: AI-generated music uploads account for <1% of total streams; AI-adoption is strongest in functional and low-stakes settings such as background music (Advanced Television)
https://www.advanced-television.com/2026/02/12/report…

What the actual fuck?
The 13yo kept wanting to know if #Mamdani has seen K-pop Demon Hunters (don't ask me why, but he was bugging me to look it up all afternoon while we were out; I was discussing Mamdani w/ the family for other reasons:

Series C, Episode 13 - Terminal
ZEN: The continuing malfunction of hull sensors is limiting data for analysis. Operational systems indicate that the damage is widespread but superficial.
DAYNA: The nature of the damage. What's causing it?
https://blake.torpidity.net/m/313/255 B7B4

Crosslisted article(s) found for cs.LG. https://arxiv.org/list/cs.LG/new
[1/3]:
- SMaRT: Online Reusable Resource Assignment and an Application to Mediation in the Kenyan Judiciary
Farabi, Pinto, Lu, Ramos-Maqueda, Das, Deeb, Sautmann
https://arxiv.org/abs/2602.18431 https://mastoxiv.page/@arXiv_csCY_bot/116119352329590193
- Benchmarking Distilled Language Models: Performance and Efficiency in Resource-Constrained Settings
Sachin Gopal Wani, Eric Page, Ajay Dholakia, David Ellison
https://arxiv.org/abs/2602.20164 https://mastoxiv.page/@arXiv_csCL_bot/116130101399805837
- VISION-ICE: Video-based Interpretation and Spatial Identification of Arrhythmia Origins via Neura...
Dorsa EPMoghaddam, Feng Gao, Drew Bernard, Kavya Sinha, Mehdi Razavi, Behnaam Aazhang
https://arxiv.org/abs/2602.20165 https://mastoxiv.page/@arXiv_csCV_bot/116130222034322594
- Benchmarking Early Deterioration Prediction Across Hospital-Rich and MCI-Like Emergency Triage Un...
KMA Solaiman, Joshua Sebastian, Karma Tobden
https://arxiv.org/abs/2602.20168 https://mastoxiv.page/@arXiv_csCY_bot/116130239074411770
- Cross-Chirality Generalization by Axial Vectors for Hetero-Chiral Protein-Peptide Interaction Design
Yang, Tian, Jia, Zhang, Zheng, Wang, Su, He, Liu, Lan
https://arxiv.org/abs/2602.20176 https://mastoxiv.page/@arXiv_qbioBM_bot/116130281674122586
- Enhancing Heat Sink Efficiency in MOSFETs using Physics Informed Neural Networks: A Systematic St...
Aniruddha Bora, Isabel K. Alvarez, Julie Chalfant, Chryssostomos Chryssostomidis
https://arxiv.org/abs/2602.20177 https://mastoxiv.page/@arXiv_csNE_bot/116130397676559696
- Data-Driven Deep MIMO Detection:Network Architectures and Generalization Analysis
Yongwei Yi, Xinping Yi, Wenjin Wang, Xiao Li, Shi Jin
https://arxiv.org/abs/2602.20178 https://mastoxiv.page/@arXiv_eessSP_bot/116130257424413457
- OrgFlow: Generative Modeling of Organic Crystal Structures from Molecular Graphs
Mohammadmahdi Vahediahmar, Matthew A. McDonald, Feng Liu
https://arxiv.org/abs/2602.20195 https://mastoxiv.page/@arXiv_condmatmtrlsci_bot/116130271189617558
- KEMP-PIP: A Feature-Fusion Based Approach for Pro-inflammatory Peptide Prediction
Soumik Deb Niloy, Md. Fahmid-Ul-Alam Juboraj, Swakkhar Shatabda
https://arxiv.org/abs/2602.20198 https://mastoxiv.page/@arXiv_qbioQM_bot/116130341315320687
- Regressor-guided Diffusion Model for De Novo Peptide Sequencing with Explicit Mass Control
Shaorong Chen, Jingbo Zhou, Jun Xia
https://arxiv.org/abs/2602.20209 https://mastoxiv.page/@arXiv_qbioQM_bot/116130374083646541
- The Sim-to-Real Gap in MRS Quantification: A Systematic Deep Learning Validation for GABA
Zien Ma, S. M. Shermer, Oktay Karaku\c{s}, Frank C. Langbein
https://arxiv.org/abs/2602.20289 https://mastoxiv.page/@arXiv_eessSP_bot/116130267228834775
- Gap-Dependent Bounds for Nearly Minimax Optimal Reinforcement Learning with Linear Function Appro...
Haochen Zhang, Zhong Zheng, Lingzhou Xue
https://arxiv.org/abs/2602.20297 https://mastoxiv.page/@arXiv_statML_bot/116130255458256497
- Multilevel Determinants of Overweight and Obesity Among U.S. Children Aged 10-17: Comparative Eva...
Joyanta Jyoti Mondal
https://arxiv.org/abs/2602.20303 https://mastoxiv.page/@arXiv_csAI_bot/116130097466859145
- An artificial intelligence framework for end-to-end rare disease phenotyping from clinical notes ...
Shyr, Hu, Tinker, Cassini, Byram, Hamid, Fabbri, Wright, Peterson, Bastarache, Xu
https://arxiv.org/abs/2602.20324 https://mastoxiv.page/@arXiv_csAI_bot/116130100089848459
- Circuit Tracing in Vision-Language Models: Understanding the Internal Mechanisms of Multimodal Th...
Jingcheng Yang, Tianhu Xiong, Shengyi Qian, Klara Nahrstedt, Mingyuan Wu
https://arxiv.org/abs/2602.20330 https://mastoxiv.page/@arXiv_csCV_bot/116130463214879334
- No One Size Fits All: QueryBandits for Hallucination Mitigation
Nicole Cho, William Watson, Alec Koppel, Sumitra Ganesh, Manuela Veloso
https://arxiv.org/abs/2602.20332 https://mastoxiv.page/@arXiv_csCL_bot/116130370809116915
- Learning During Detection: Continual Learning for Neural OFDM Receivers via DMRS
Mohanad Obeed, Ming Jian
https://arxiv.org/abs/2602.20361 https://mastoxiv.page/@arXiv_csIT_bot/116130289537785136
- Detecting and Mitigating Group Bias in Heterogeneous Treatment Effects
Joel Persson, Jurri\"en Bakker, Dennis Bohle, Stefan Feuerriegel, Florian von Wangenheim
https://arxiv.org/abs/2602.20383 https://mastoxiv.page/@arXiv_statME_bot/116130509065601748
- Selecting Optimal Variable Order in Autoregressive Ising Models
Shiba Biswal, Marc Vuffray, Andrey Y. Lokhov
https://arxiv.org/abs/2602.20394 https://mastoxiv.page/@arXiv_statML_bot/116130299369541741
toXiv_bot_toot

Fuck, no, Apple. I was forced to update to iOS 26 due to lack of security updates for 18 on devices that can run 26 (i.e., due to extortion) but you’re not infecting my Mac with Liquid Ass.
Notice the dark pattern of pestering you in hopes you will accidentally hit the “Restart” button even though you’ve specifically turned off auto updates. Microsoft-level disrespect for consent here.
(Not surprising from a company that gives golden awards to fascists.)

Where's the fun in using a mid to high [and at that, an AMD] GPU for Linux gaming? 🙂 Even now, most gamers are broke by default, so I am using my second budget build that I had set aside for Win11 to do my Windows chores, but have completely migrated to Linux now, to showcase how real™ men™ overcome difficulties and optimise settings even on garbage PCs 😃
Linux Budget Gaming - Detroit: Become Human - GTX 1650 GDDR6 4GB ( i5-13400F)

Brainwaves meet spatial computing: Cognixion Vision Pro #noninvasive

The Diffusion Duality, Chapter II: $\Psi$-Samplers and Efficient Curriculum
Justin Deschenaux, Caglar Gulcehre, Subham Sekhar Sahoo
https://arxiv.org/abs/2602.21185 https://arxiv.org/pdf/2602.21185 https://arxiv.org/html/2602.21185
arXiv:2602.21185v1 Announce Type: new
Abstract: Uniform-state discrete diffusion models excel at few-step generation and guidance due to their ability to self-correct, making them preferred over autoregressive or Masked diffusion models in these settings. However, their sampling quality plateaus with ancestral samplers as the number of steps increases. We introduce a family of Predictor-Corrector (PC) samplers for discrete diffusion that generalize prior methods and apply to arbitrary noise processes. When paired with uniform-state diffusion, our samplers outperform ancestral sampling on both language and image modeling, achieving lower generative perplexity at matched unigram entropy on OpenWebText and better FID/IS scores on CIFAR10. Crucially, unlike conventional samplers, our PC methods continue to improve with more sampling steps. Taken together, these findings call into question the assumption that Masked diffusion is the inevitable future of diffusion-based language modeling. Beyond sampling, we develop a memory-efficient curriculum for the Gaussian relaxation training phase, reducing training time by 25% and memory by 33% compared to Duo while maintaining comparable perplexity on OpenWebText and LM1B and strong downstream performance. We release code, checkpoints, and a video-tutorial on: https://s-sahoo.com/duo-ch2
toXiv_bot_toot

🥳 JavaScript Database (JSDB)¹ version 7.0.0 released
- *Breaking change* JSTable.PERSIST event now uses a parameter object with properties for `type`, `keypath`, `value`, `change`, and `table`. This should make listening for events on your databases much nicer to author. e.g., a snippet from Catalyst² I’m working on:
```js
const settingsTable = db.settings['__table__']
const JSTable = settingsTable.constructor
settingsTable.addListener(JSTable.PERSIST, (…

Deep unfolding of MCMC kernels: scalable, modular & explainable GANs for high-dimensional posterior sampling
Jonathan Spence, Tob\'ias I. Liaudat, Konstantinos Zygalakis, Marcelo Pereyra
https://arxiv.org/abs/2602.20758 https://arxiv.org/pdf/2602.20758 https://arxiv.org/html/2602.20758
arXiv:2602.20758v1 Announce Type: new
Abstract: Markov chain Monte Carlo (MCMC) methods are fundamental to Bayesian computation, but can be computationally intensive, especially in high-dimensional settings. Push-forward generative models, such as generative adversarial networks (GANs), variational auto-encoders and normalising flows offer a computationally efficient alternative for posterior sampling. However, push-forward models are opaque as they lack the modularity of Bayes Theorem, leading to poor generalisation with respect to changes in the likelihood function. In this work, we introduce a novel approach to GAN architecture design by applying deep unfolding to Langevin MCMC algorithms. This paradigm maps fixed-step iterative algorithms onto modular neural networks, yielding architectures that are both flexible and amenable to interpretation. Crucially, our design allows key model parameters to be specified at inference time, offering robustness to changes in the likelihood parameters. We train these unfolded samplers end-to-end using a supervised regularized Wasserstein GAN framework for posterior sampling. Through extensive Bayesian imaging experiments, we demonstrate that our proposed approach achieves high sampling accuracy and excellent computational efficiency, while retaining the physics consistency, adaptability and interpretability of classical MCMC strategies.
toXiv_bot_toot

On the Generalization Behavior of Deep Residual Networks From a Dynamical System Perspective
Jinshu Huang, Mingfei Sun, Chunlin Wu
https://arxiv.org/abs/2602.20921 https://arxiv.org/pdf/2602.20921 https://arxiv.org/html/2602.20921
arXiv:2602.20921v1 Announce Type: new
Abstract: Deep neural networks (DNNs) have significantly advanced machine learning, with model depth playing a central role in their successes. The dynamical system modeling approach has recently emerged as a powerful framework, offering new mathematical insights into the structure and learning behavior of DNNs. In this work, we establish generalization error bounds for both discrete- and continuous-time residual networks (ResNets) by combining Rademacher complexity, flow maps of dynamical systems, and the convergence behavior of ResNets in the deep-layer limit. The resulting bounds are of order $O(1/\sqrt{S})$ with respect to the number of training samples $S$, and include a structure-dependent negative term, yielding depth-uniform and asymptotic generalization bounds under milder assumptions. These findings provide a unified understanding of generalization across both discrete- and continuous-time ResNets, helping to close the gap in both the order of sample complexity and assumptions between the discrete- and continuous-time settings.
toXiv_bot_toot