Tootfinder

14/ Prominenz aus Funk und Fernsehen war auch da. Ich habe @… von der #LetzteGeneration gefragt, ob ich sie fotografieren könnte und sie meinte, so direkt von vorne ginge aber nicht. Das wirke total gestellt. Hab ich eingesehen, aber wenn ich sie fotografiere, wie sie mir erklärt, dass direkt von vorne nicht ginge und gestellt wirke, dann ist das OK, oder?
Tim Wechselmann-Cassim meinte: Hey, Du bist doch der, der damals in der Hasenheide die Portraits gemacht hat! Wir müssten das jetzt noch mal machen, damit man dann mal sehen kann, was zwei Jahre Aktivismus mit einem so machen.
Und ja, daran musste ich heute denken, als ich den Artikel über #KI-generierte Bilder in der taz las: „Lange wird das allerdings nicht so bleiben. Während die KI-Modelle zunehmend besser werden, experimentieren Nutzer*innen in Foren mit Prompts und Zusatzprogrammen, mit Kontrollnetzwerken und Anpassungsmodellen. Sie haben ein großes gemeinsames Ziel: Aufwendig bügeln sie Unregelmäßigkeiten in die digital-glatte Haut hinein – für ein bisschen Charakter.“
#CarlaHinrichs

Tschechien: Geld für 300.000 Schuss Artilleriemunition gesammelt
Die von Tschechien geführte internationale Initiative für den Einkauf von Waffen für die Ukraine außerhalb von Europa hat nach Angaben von Ministerpräsident Petr Fiala Geld für rund 300.000 Schuss Artilleriemunition gesammelt. Ziel sei es jedoch, "noch viel mehr zu leisten", erklärte Fiala im Onlinedienst X. "Wir suchen weiterhin Partne…
🐦

Petr Fiala (@P_Fiala) on X
Many thanks to all the countries that have joined the Czech initiative to purchase ammunition for Ukraine so far. We have managed to raise enough money to buy the first batch of 300,000 artillery shells. However, our goal is to deliver much more! Our work and our help to Ukraine…

This https://arxiv.org/abs/2402.19379 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCY_…

Wisdom of the Silicon Crowd: LLM Ensemble Prediction Capabilities Match Human Crowd Accuracy
Human forecasting accuracy in practice relies on the 'wisdom of the crowd' effect, in which predictions about future events are significantly improved by aggregating across a crowd of individual forecasters. Past work on the forecasting ability of large language models (LLMs) suggests that frontier LLMs, as individual forecasters, underperform compared to the gold standard of a human crowd forecasting tournament aggregate. In Study 1, we expand this research by using an LLM ensemble approach co…

JAX-SPH: A Differentiable Smoothed Particle Hydrodynamics Framework
Artur P. Toshev, Harish Ramachandran, Jonas A. Erbesdobler, Gianluca Galletti, Johannes Brandstetter, Nikolaus A. Adams
https://arxiv.org/abs/2403.04750

JAX-SPH: A Differentiable Smoothed Particle Hydrodynamics Framework
Particle-based fluid simulations have emerged as a powerful tool for solving the Navier-Stokes equations, especially in cases that include intricate physics and free surfaces. The recent addition of machine learning methods to the toolbox for solving such problems is pushing the boundary of the quality vs. speed tradeoff of such numerical simulations. In this work, we lead the way to Lagrangian fluid simulators compatible with deep learning frameworks, and propose JAX-SPH - a Smoothed Particle …

I am now a Ko-fi Ambassador! 🥳
#Kofi

Support Andy Carolan on Ko-fi! ❤️
I am using Ko-fi to offer cute profile picture commissions, icons, headers, backgrounds, zines, and other neat things. Also, if you like my work, and want to support me, please consider buying me a Flat White!

There are 14 people on the W3C's Social Web Incubator Community Group today. The group is finalizing new documents that describe how to use WebFinger and HTTP Signatures in the #Fediverse, and a bunch of other work going on as well.
So glad to see what there is ongoing work making the technical plumbing that makes the Fediverse work even better!
If you care, you can participate, it…

Social Web Incubator Community Group
The Social Web Incubator Community Group (also known as SocialCG, or SWICG) is the successor of the Social Web Working Group, which ran from 2014 to 2017. The SocialCG provides space to collaborate and coordinate for implementors who are building on any of the specifications published by the Social Web WG, and related technologies. It is also a place to incubate new proposals which build on or complement the Social Web WG recommendations…

RankSHAP: a Gold Standard Feature Attribution Method for the Ranking Task
Tanya Chowdhury, Yair Zick, James Allan
https://arxiv.org/abs/2405.01848 https://…

RankSHAP: a Gold Standard Feature Attribution Method for the Ranking Task
Several works propose various post-hoc, model-agnostic explanations for the task of ranking, i.e. the task of ordering a set of documents, via feature attribution methods. However, these attributions are seen to weakly correlate and sometimes contradict each other. In classification/regression, several works focus on \emph{axiomatic characterization} of feature attribution methods, showing that a certain method uniquely satisfies a set of desirable properties. However, no such efforts have been…

The Armitage Account - Season 1 finale 🐙📚🔥
Dramatized horror anthology | Total listening time: 3h 35m
Isaac Andrews, plagued with guilt over the events of the last year, begins to search for a connection to his supposed great-grandfather, and the long-deceased head librarian of Miskatonic University, Henry Armitage. Inspired by the works...

The Armitage Account - Season 1 finale 🐙📚🔥
Dramatized horror anthology | Total listening time: 3h 35m Isaac Andrews, plagued with guilt over the events of the last year, begins to search for a connection to his supposed great-grandfather, and the long-deceased head librarian of Miskatonic University, Henry Armitage. Inspired by the works of H.P. Lovecraft. Produced with a companion ARG that has yet to be solved by the fandom. 1 seasons, 7 episodes as of Feb 21, 2024.

This https://arxiv.org/abs/2402.19379 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCY_…

Wisdom of the Silicon Crowd: LLM Ensemble Prediction Capabilities Rival Human Crowd Accuracy
Human forecasting accuracy in practice relies on the 'wisdom of the crowd' effect, in which predictions about future events are significantly improved by aggregating across a crowd of individual forecasters. Past work on the forecasting ability of large language models (LLMs) suggests that frontier LLMs, as individual forecasters, underperform compared to the gold standard of a human crowd forecasting tournament aggregate. In Study 1, we expand this research by using an LLM ensemble approach co…