iiinova — where your digital self keeps evolving

Reposted by Daniel Augusto

Daniel Rothenberg @danielrothenberg.com · 11/08/2026

How good are the frontier AI weather forecast models from Google DeepMind, Microsoft, and NVIDIA? See for yourself on @brightband.com's Operational WeatherBench dashboard - owb.brightband.com! We're happy to run these models operationally and share our evaluation stats with the community!

owb.brightband.com

Operational WeatherBench | AI Weather Forecast Leaderboard

Live skill rankings for AI and physics-based weather forecast models, scored daily with standard metrics and an open toolkit.

0134

Reposted by Daniel Augusto

残響室 @45978.bsky.social · 30/05/2026

今これ

13712

Reposted by Daniel Augusto

Rui-Yang Zhang @ryzhang.bsky.social · 21/05/2026

Henry and co-authors arxived a really exciting paper today (arxiv.org/abs/2605.21041) on a new paradigm for Gaussian Process regression where you can condition on (almost) ANYTHING, even LLM prompts!

0175

Reposted by Daniel Augusto

Mark Riedl @markriedl.bsky.social · 14/05/2026

ArXiV has a new LLM policy (Screenshots with alt text so you don’t have to click through to the other place and see all the stupid responses)

Attention @arxiv authors: Our Code of Conduct states that by signing your name as an author of a paper, each author takes full responsibility for all its contents, irrespective of how the contents were generated. 1/

If generative AI tools generate inappropriate language, plagiarized content, biased content, errors, mistakes, incorrect references, or misleading content, and that output is included in scientific works, it is the responsibility of the author(s). 2/

We have recently clarified our penalties for this. If a submission contains incontrovertible evidence that the authors did not check the results of LLM generation, this means we can't trust anything in the paper. 3/

The penalty is a 1-year ban from arXiv followed by the requirement that subsequent arXiv submissions must first be accepted at a reputable peer-reviewed venue. 4/

Examples of incontrovertible evidence: hallucinated references, meta-comments from the LLM ("here is a 200 word summary; would you like me to make any changes?"; "the data in this table is illustrative, fill it in with the real numbers from your experiments") end/

8312105

Reposted by Daniel Augusto

Sam Power @spmontecarlo.bsky.social · 15/05/2026

Always quite jarring for me to see discussions about arXiv which reveal how many people see the moderation as gatekeeping, or otherwise negative. Muddies the waters for me as to what they think the point of the website is.

2314

Reposted by Daniel Augusto

Phil Lewis @phillewis.bsky.social · 15/05/2026

Dozens of empty Waymos invaded an Atlanta neighborhood and circled a cul-de-sac for hours with no passengers www.wsbtv.com/news/local/a...

58582712267

Reposted by Daniel Augusto

Georg Bökman @bokmangeorg.bsky.social · 31/03/2026

In 1994 Fefferman proved that by analytically continuing a neural network with sigmoid-activations to the complex numbers, you can read off the weights of the network by looking at the pole structure of the function. www.math.stonybrook.edu/~bishop/clas...

1132

Reposted by Daniel Augusto

Stephen Bush @stephenkb.bsky.social · 07/05/2026

This terrific London Centric story is I think illustrative: whenever someone actually looks into phone theft, huh, what do you know, it's organised crime, it's not opportunists looking to make a buck.

londoncentric.media

London's thieves are burying phones in flowerbeds

The police don't take much interest in returning the stolen devices when they're dug up — so London Centric decided to do it.

928039

Reposted by Daniel Augusto

derek guy @dieworkwear.bsky.social · 05/05/2026

Pierre Bourdieu once argued that the question of whether something is art isn't answered by the artist's intention or even the object itself. Instead, the quality of being "art" is conferred by a community — curators, dealers, auction houses, audiences, etc. 🧵

Someone on Threads writes: "The Met Gala feels like real dystopian shit.

People can’t afford groceries or gas. But we’re going to celebrate this event which is nothing more than wealth porn."

18079711723

Reposted by Daniel Augusto

Vincent D. Warmerdam @koaning.bsky.social · 01/05/2026

People should not sleep on the new EVoC clustering algorithm. It's got me rethinking a lot of my widgets and bulk labelling tricks. youtu.be/1roN51JXuX8

youtube.com

marimo

The Most Satisfying Method to Explore Embeddings

1155

Reposted by Daniel Augusto

Joachim Baumann @joachimbaumann.bsky.social · 01/05/2026

Can you boost your AI review scores by asking an LLM to rewrite your paper? Yes! We call it paper laundering Our @icmlconf.bsky.social spotlight paper argues current AI reviewers aren't ready to automate peer review, and outlines what a science of peer review automation should look like 🧵👇 #ICML2026

First page of the ICML 2026 spotlight paper "Stop Automating Peer Review Without Rigorous Evaluation" by Joachim Baumann, Jiaxin Pei, Sanmi Koyejo, and Dirk Hovy (Stanford University and Bocconi University). The abstract argues that today's AI systems should not be used to produce paper reviews, grounded in two empirical findings: a "hivemind effect" where AI reviewers show excessive agreement and reduce perspective diversity, and "paper laundering," where prompting an LLM to rewrite a paper trivially increases AI reviewer scores through stylistic changes rather than scientific improvements. The paper calls for a science of peer review automation rather than wholesale deployment of general-purpose LLMs.

44314

Reposted by Daniel Augusto

Firefox for Web Developers @webdevs.firefox.com · 30/04/2026

Chrome looks set to ship an LLM Prompt API to the web platform. At Mozilla, we oppose this API. We feel it has a large interoperability risk, and Google imposing T&Cs on a web API sets a dangerous precedent. Full details: github.com/mozilla/stan...

github.com

Prompt API · Issue #1213 · mozilla/standards-positions

Specification title Prompt API Specification or proposal URL (if available) No response Explainer URL (if available) https://github.com/webmachinelearning/prompt-api/blob/main/README.md Proposal au...

1139697

Reposted by Daniel Augusto

cabanier.bsky.social @cabanier.bsky.social · 23/04/2026

Quest Browser 146.0 now has experimental support for WebGPU in #WebXR. Turn on "webXR experimental features" in "chrome://flags" and the API will be available. Thanks @toji.dev for the demo file! Feedback welcome.

0166

Reposted by Daniel Augusto

VIEWS project @viewsproject.bsky.social · 20/04/2026

Roll up for the VIEWS summer seminars! We have Egyptian graffiti, the Phrygian alphabet, early cuneiform animals and the ultimate question of how writing grows up and develops... Two hybrid sessions, two online only. Join us! viewsproject.wordpress.com/2026/04/20/v...

viewsproject.wordpress.com

VIEWS Summer seminars 2026

We are delighted to announce an exciting programme of seminars for the coming term, taking place in May and June 2026. The first two seminars are in-person/hybrid, the second two online only &#8211…

286

Reposted by Daniel Augusto

Yining Karl Li @yiningkarlli.bsky.social · 21/04/2026

Disney Animation remade three songs for ASL. For these, we resurrected the original data and pipeline version, reanimated (from near scratch in some places), relit where needed, and re-rendered. Probably the hardest way to do this, but definitely the right way to do it, so that's what we did!

212333

Reposted by Daniel Augusto

Nikita Lisitsa @lisyarus.bsky.social · 14/04/2026

New blog post! I really hope I'll get a decent amount of people mad with this one 😈😈😈 #programming #computing #floatingpoint #geometry lisyarus.github.io/blog/posts/i...

lisyarus.github.io

It's OK to compare floating-points for equality

1010421

Reposted by Daniel Augusto

Marc Deisenroth @deisenroth.bsky.social · 12/04/2026

If you are planning to submit a paper to @neuripsconf.bsky.social and you need to create a new @openreview.bsky.social profile, do it today: OpenReview's profile moderation process is stricter than before, thus it may take longer for new users to have their account activated (up to 2 weeks).

0177

Reposted by Daniel Augusto

uai2026 @auai.org · 10/04/2026

📣 Registration for #uai2026 is now open! Join us in Amsterdam 🇳🇱 this August 17-21 for cutting-edge research on uncertainty in AI. early-bird rates available 🐦 students welcome 🎓 Secure your spot 👉 auai.org/uai2026/registration

auai.org

Uncertainty in Artificial Intelligence

Amsterdam, Netherlands | 17-21 August

0126

Reposted by Daniel Augusto

Kelly² @anarchakelly.bsky.social · 10/04/2026

You should turn off message content in your notification settings. No real point in using an end to end encrypted app if people can just read the push notification.

11838411

Reposted by Daniel Augusto

David Picard @davidpicard.eurosky.social · 08/04/2026

🚨 arxiv.org/abs/2604.06129 PoM: A Linear-Time Replacement for Attention with the Polynomial Mixer This paper is the result of doing a lab-wide hackathon on an idea I've had for some time. Probably the paper with the highest number of authors I've ever done. It's a CVPR Findings 26. Thread 🧵👇

arxiv.org

PoM: A Linear-Time Replacement for Attention with the Polynomial Mixer

This paper introduces the Polynomial Mixer (PoM), a novel token mixing mechanism with linear complexity that serves as a drop-in replacement for self-attention. PoM aggregates input tokens into a comp...

46520

Reposted by Daniel Augusto

Katie Mack @astrokatie.com · 03/04/2026

More context on this #Artemis II image: * This is the night side, lit by moonlight. You can see city lights in Spain & Portugal, & a sliver of day at lower right * The Sun is entirely behind Earth, which makes it a kind of solar eclipse, but w/ Earth doing the eclipsing instead of the Moon: ☀️🌍🚀🌕

A full disc image of Earth, as seen from the Orion Crew Module. The planet is a pale blue, swirling with white clouds and glowing slightly lighter blue in place from reflected light. At lower left, a large brown landmass is Africa, with Spain and Portugal with twinkling lights where the planet curves. At top right, auroras glow in a thin green glow, just barely separated from the planet's surface. Earth is set against the black of space (pic: NASA/R.Wiseman)

231130443678

Reposted by Daniel Augusto

Nathan Lambert @natolambert.bsky.social · 02/04/2026

Google dropped 4 different Gemma open-weight models! I'm most excited that they're finally adopting a standard Apache 2.0 open source license. huggingface.co/collections/...

211021

Reposted by Daniel Augusto

Peter Henderson @peterhenderson.bsky.social · 27/03/2026

This is a challenging legal problem for NeurIPS (and other conference participants)! You might be wondering how this is possible given the First Amendment? I wrote a quick explainer on the current status quo of relevant First Amendment cases & law to get you up to speed. 🔗👇

185

Reposted by Daniel Augusto

Savannah Ostrowski @savannah.dev · 25/03/2026

God, I really hate what open source is becoming. Commenting a humongous block of text as an "analysis" with a line at the end that you are an AI agent is not helpful. Not to be dramatic but the volume of slop right now is an attack on maintainers. github.com/python/cpyth...

github.com

Improving trace quality by tracking "fitness" and "exit quality" · Issue #146073 · python/cpython

Feature or enhancement Proposal: The performance of the JIT depends heavily on the quality of the traces that the front-end produces. If traces are too short, very long, overlap a lot, or drop into...

67712

Reposted by Daniel Augusto

NeurIPS Europe @neuripseurope.bsky.social · 23/03/2026

Today NeurIPS is announcing our official satellite event in Paris. After responding to the call from Ellis following the success of EurIPS in December, we are pleased to reach a new milestone by joining forces with the NeurIPS organizing committee for the 2026 edition.

18933

Reposted by Daniel Augusto

Dan Conway @magisterconway.bsky.social · 22/03/2026

So, um... this is bad. Really bad. I looked at the letters that were translated by the AI, and the very first one I found was almost entirely hallucination. Thread:

3639212001

Reposted by Daniel Augusto

Marc Deisenroth @deisenroth.bsky.social · 16/03/2026

@neuripsconf.bsky.social call for papers is out: neurips.cc/Conferences/... Key dates: - Abstract submission deadline: 4 May (AOE). - Full paper submission deadline (incl. all supplementary materials): 6 May - Author notification: 24 September

neurips.cc

Call for Papers 2026

074

Reposted by Daniel Augusto

garrison @garrison.corporate.fm · 15/03/2026

nano banana infographics must be stopped at any cost

3739

Reposted by Daniel Augusto

Eurosky Social @eurosky.social · 15/12/2025

Announcing eurosky.social accounts - launching January 2026. ✅ Managed by Eurosky, a European non-profit initiative ✅ Hosted on European cloud ✅ Governed by European law www.eurosky.social/register

eurosky.social

501189640

Reposted by Daniel Augusto

Viacheslav Borovitskiy @vabor112.bsky.social · 12/03/2026

Happy to share a major milestone: after years of development, we are officially launching Version 1.0 of the GeometricKernels library! To top it off, our accompanying paper has just been published in JMLR (MLOSS)! 🎉 github.com/geometric-ke...

14812

Reposted by Daniel Augusto

シャポコ🌵 @shapoco.net · 26/02/2026

|￣￣￣￣￣￣￣￣￣￣￣￣￣￣￣| |　　　　　　　　　　　　　　　| |　　　　　　　　　　　　　　　| |　　　　　　　　　　　　　　　| |　　　　　　　　　　　　　　　| |　　　　　　　　　　　　　　　| |　　　　　　　　　　　　　　　| |　　𝑺𝑬𝑬 𝒀𝑶𝑼 𝑺𝑷𝑨𝑪𝑬 𝑪𝑶𝑾𝑩𝑶𝒀...| ￣￣￣￣￣￣￣￣￣￣￣￣￣￣￣

0246

Reposted by Daniel Augusto

shar @sharyap.com · 21/02/2026

happy 40th zelda!!! ⚔🛡 hyah

112398628

Reposted by Daniel Augusto

David Buchanan @retr0.id · 15/02/2026

echo "set -g mouse on" >> ~/.tmux.conf

3311

Reposted by Daniel Augusto

Martin Trapp @trappmartin.eurosky.social · 21/01/2026

Accurate and thorough representation of prior and related work is one of the cornerstones of good research. It is shocking to me that so many published NeurIPS papers, even from top institutions, have fabricated references. I recommend reading the original report: gptzero.me/news/neurips/

gptzero.me

GPTZero finds 100 new hallucinations in NeurIPS 2025 accepted papers

GPTZero's analysis 4841 papers accepted by NeurIPS 2025 show there are at least 100 with confirmed hallucinations

1339

Reposted by Daniel Augusto

Mike Sux 🍁 @distraction.engineer · 04/01/2026

"Wide Event" logging for better debugging. loggingsucks.com Video summary with Theo: youtu.be/wFYdok_YWq8?...

loggingsucks.com

Logging Sucks - Your Logs Are Lying To You

Why traditional logging fails and how wide events can fix your observability

092

Daniel Augusto @spectral.space · 30/12/2025

Honestly, I just want a feed that normalizes the time zones of who I'm following 🥲 I dislike the random recommendations of network-based feeds but they are the best way to see what happened when I was sleeping without having to scroll tons

000