Tootfinder

Source: Disney saw more than 1.7M total paid streaming cancelations between September 17, when it suspended Jimmy Kimmel, and September 23 (Marisa Kabas/@marisakabas.bsky.social)
https://bsky.app/profile/marisakabas.bsky.social/post/3lzyjg3ulo222

Marisa Kabas (@marisakabas.bsky.social)
SCOOP / UPDATE — Disney saw more than 1.7 million total paid streaming cancelations during the period 9/17-9/23, a Disney source confirms to me. The total includes Disney+, Hulu and ESPN. [contains quote post or other embedded content]

Seismic evidence for a highly heterogeneous martian mantle: #Marsquake Data Reveals Lumpy Nature of Red Planet’s Interior: https://www.nasa.gov/missions/insight/nasa-marsquake-data-reveals-lumpy-nature-of-red-planets-interior/

Can't-Miss Play: Markquese Bell runs over 80 YARDS to take blocked extra point back for a deuce https://www.nfl.com/videos/can-t-miss-play-markquese-bell-runs-over-80-yards-to-take-blocked-extra-point-back-for-a-deuce

»Margaux Blanchard – US-Medien fallen auf KI-generierte Artikel von Freelancerin herein:
Mehrere US-Medien haben offenbar Artikel einer Freelancerin namens Margaux Blanchard veröffentlicht, die komplett per KI erstellt worden sein sollen. Die Artikel sind mittlerweile entfernt. Aber eine Frage bleibt: Wie konnte das passieren?«
Die Frage ist doch, wie überaschend ist dies und wiso werden jenste plumpe News geteilt?
🤖

Margaux Blanchard: US-Medien fallen auf KI-generierte Artikel von Freelancerin herein | t3n
Mehrere US-Medien haben offenbar Artikel einer Freelancerin namens Margaux Blanchard veröffentlicht, die komplett per KI erstellt worden sein sollen. Die Artikel sind mittlerweile entfernt. Aber eine Frage bleibt: Wie konnte das passieren? Im Mai 2025 war die Chicago Sun-Times von einem KI-Debakel durchgerüttelt worden, als in ihrer Sonntagsbeilage Lektüretipps und Empfehlungen zu Sommeraktivitäten erschienen, die […]

MARCOS: Deep Thinking by Markov Chain of Continuous Thoughts
Jiayu Liu, Zhenya Huang, Anya Sims, Enhong Chen, Yee Whye Teh, Ning Miao
https://arxiv.org/abs/2509.25020 https://…

MARCOS: Deep Thinking by Markov Chain of Continuous Thoughts
The current paradigm for reasoning in large language models (LLMs) involves models "thinking out loud" via a sequence of tokens, known as chain-of-thought (CoT). This approach, while effective, has several significant drawbacks. Firstly, inference requires autoregressive generation of often thousands of CoT tokens, which is slow and computationally expensive. Secondly, it constrains reasoning to the discrete space of tokens, creating an information bottleneck across reasoning steps. Thirdly, it…

Simple in-place yet comparison-optimal Mergesort
Christian Siebert
https://arxiv.org/abs/2509.24540 https://arxiv.org/pdf/2509.24540

Simple in-place yet comparison-optimal Mergesort
Mergesort is one of the few efficient sorting algorithms and, despite being the oldest one, often still the method of choice today. In contrast to some alternative algorithms, it always runs efficiently using O(n log n) element comparisons and usually works in a stable manner. Its only practical disadvantage is the need for a second array, and thus twice the amount of memory. This can be an impeding factor, especially when handling large amounts of data, where it is often either impractical or …

Ashton Jeanty joins Raider greats Bo Jackson and Marcus Allen with feat, sets other franchise, NFL milestones

https://www.cbssports.com/nfl/news/ashton-

How Brazil's Pix digital payment system, adopted by 80% of citizens, became a target of the Trump administration, which says it unfairly undercuts US companies (Ana Ionova/New York Times)
https://www.nytimes.com/2025/09/29/world/americas/brazil-digital-pay…

Brazil’s Homegrown Payment System Is Target of Trump Administration
Brazil’s fast and free homegrown digital payment system, PIX, has become wildly popular. The Trump administration says it unfairly undercuts U.S. companies.

GaLactic and extragalactic all-sky Murchison Widefield Array survey eXtended (GLEAM-X) - Galactic plane: #MilkyWay reveals our Galaxy in unprecedented radio colour: https://www.icrar.org/gleam-x-galactic-plane/ - astronomers from ICRAR have created the largest low-frequency radio colour image of the Milky Way ever assembled -> New images reveal the Milky Way’s stunning galactic plane in more detail than ever before: https://theconversation.com/new-images-reveal-the-milky-ways-stunning-galactic-plane-in-more-detail-than-ever-before-264688

Rethinking Layer-wise Model Merging through Chain of Merges
Pietro Buzzega, Riccardo Salami, Angelo Porrello, Simone Calderara
https://arxiv.org/abs/2508.21421 https://

Rethinking Layer-wise Model Merging through Chain of Merges
Fine-tuning pretrained models has become a standard pathway to achieve state-of-the-art performance across a wide range of domains, leading to a proliferation of task-specific model variants. As the number of such specialized modules in-creases, merging them into a unified model without retraining has become a critical challenge. Existing merging techniques often rely on interference heuristics,importance weighting, or activation matching while treating each layer independently, thereby failing…