Tootfinder

Y'all were wondering which way the army would break. This is as clear an answer as you're going to get.
This isn't a mistake. It's a message.
Edit: cut out "During drills they were perfectly in time" because I don't have receipts to show. Linking to a longer video in a threads post that shows a bit more intense of a juxtaposition between themselves and the division in front of them. This thread makes the same claim about drills, but I also don't see video.
https://www.threads.com/@davidmorehouse/post/DK7ExrKs5iw?xmt=AQF0w52awa66lUQeIuolzDaz8uSFcMmJLHALm9RZc4P75w
Edit: to clarify a bit, this is one division. The division in front of them in the longer video is in step. This isn't "the whole army is going to refuse" but this may be, "some units are fed up enough to make it clear they don't want to play, which is saying a lot."
Any division that exists needs to be cultivated. The way you cultivate that is absolutely not by lumping them all together. Some of the military voted for Trump, some don't care, some hate him. All of them are being treated like shit right now. All of them swore an oath to defend the constitution. Some of them probably know what that means.
I'm absolutely not a fan of the military, but I will take every opportunity I can to humanize these folks. Look for every opportunity you can to remind them that they are welcome on our side whenever they're ready to actually fulfill their oath, and we will appreciate every subtle bit of resistance they give to those who are violating it in the meantime.

Two-Timescale Sum-Rate Maximization for Movable Antenna Enhanced Systems
Xintai Chen, Biqian Feng, Yongpeng Wu, Derrick Wing Kwan Ng, Robert Schober
https://arxiv.org/abs/2509.04062

Two-Timescale Sum-Rate Maximization for Movable Antenna Enhanced Systems
This paper studies a novel movable antenna (MA)-enhanced multiuser multiple-input multiple-output downlink system designed to improve wireless communication performance. We aim to maximize the average achievable sum rate through two-timescale optimization exploiting instantaneous channel state information at the receiver (I-CSIR) for receive antenna position vector (APV) design and statistical channel state information at the transmitter (S-CSIT) for transmit APV and covariance matrix design. W…

Low-resource keyword spotting using contrastively trained transformer acoustic word embeddings
Julian Herreilers, Christiaan Jacobs, Thomas Niesler
https://arxiv.org/abs/2506.17690

Low-resource keyword spotting using contrastively trained transformer acoustic word embeddings
We introduce a new approach, the ContrastiveTransformer, that produces acoustic word embeddings (AWEs) for the purpose of very low-resource keyword spotting. The ContrastiveTransformer, an encoder-only model, directly optimises the embedding space using normalised temperature-scaled cross entropy (NT-Xent) loss. We use this model to perform keyword spotting for radio broadcasts in Luganda and Bambara, the latter a severely under-resourced language. We compare our model to various existing AWE a…