Tootfinder

Last season’s updated vaccine provides significant protection against covid-19,
particularly against severe illness and death.
The results reaffirm the benefit of staying current with covid vaccination.

Researchers analyzed data from approximately 1.8 million Americans between August 2024 and April 2025,
of whom only about 13 percent received the 2024–2025 coronavirus shot.
They evaluated three clinical outcomes related to the disease: infections, emergency departm…

How long does covid booster protection last? A new study offers answers.
More evidence highlighting the benefit, and limitations, of covid-19 vaccines

Dmitri #Peskow, Leiter der Russischen Präsidialverwaltung, kommentiert Darth #Trump|s Angriff auf die EU und Anbiederung an #Russland im ZDF Beitrag so: "Die neue US Außenpolitik stimmt mit unseren Vision überein.&…

Revisiting Mixout: An Overlooked Path to Robust Finetuning
Masih Aminbeidokhti, Heitor Rapela Medeiros, Eric Granger, Marco Pedersoli
https://arxiv.org/abs/2510.06982 https://…

Revisiting Mixout: An Overlooked Path to Robust Finetuning
Finetuning vision foundation models often improves in-domain accuracy but comes at the cost of robustness under distribution shift. We revisit Mixout, a stochastic regularizer that intermittently replaces finetuned weights with their pretrained reference, through the lens of a single-run, weight-sharing implicit ensemble. This perspective reveals three key levers that govern robustness: the \emph{masking anchor}, \emph{resampling frequency}, and \emph{mask sparsity}. Guided by this analysis, we…

Brennpunkt: US-Angriff auf Venezuela - #ARD #swr
https://www.ardmediathek.de/video/brennpunkt/us-angriff-auf-venezuela-oder-brennpunkt/das-erste/Y3JpZDovL3dkci5kZS9CZWl0cmFnLXNvcGhvcmEtNDk4MGIyODMtMWMyZS00ODcxLWE2MmUtMmNlNDU3ZmMwYjI0 (15min)
Op Das Erste voegt Brennpunt een heel overzicht toe van eerdere ingrepen vsn de VS in de achtertuin van de VS.

“Well just be vegan!”
Please note that I’m not going to grace any such reply with an answer, however you will get blocked.

🔄 Seeing double: Visual anagrams that rotate open doors for brain research
#vision

Seeing double: Visual anagrams that rotate open doors for brain research
New artificial intelligence-generated images that appear to be one thing, but something else entirely when rotated, are helping scientists test the human mind.

Hallucination Filtering in Radiology Vision-Language Models Using Discrete Semantic Entropy
Patrick Wienholt, Sophie Caselitz, Robert Siepmann, Philipp Bruners, Keno Bressem, Christiane Kuhl, Jakob Nikolas Kather, Sven Nebelung, Daniel Truhn
https://arxiv.org/abs/2510.09256

Hallucination Filtering in Radiology Vision-Language Models Using Discrete Semantic Entropy
To determine whether using discrete semantic entropy (DSE) to reject questions likely to generate hallucinations can improve the accuracy of black-box vision-language models (VLMs) in radiologic image based visual question answering (VQA). This retrospective study evaluated DSE using two publicly available, de-identified datasets: (i) the VQA-Med 2019 benchmark (500 images with clinical questions and short-text answers) and (ii) a diagnostic radiology dataset (206 cases: 60 computed tomography …

Angered by the Democratic leadership’s fecklessness and lack of a bold vision for the future,
a group of senators including Bernie Sanders of Vermont and Elizabeth Warren of Massachusetts
has formed an alliance to push back on Senate Minority Leader Chuck Schumer and the party’s campaign arm ahead of next year’s critical midterm elections.
The “Fight Club” is pressing the Democratic Party to
“embrace candidates willing to challenge entrenched corporate interests,

Sanders, Warren Help Form Senate Democratic ‘Fight Club’ Challenging Schumer’s Leadership | Common Dreams
"So glad there are some Senate Dems willing to fight back," said one progressive strategist.

Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space
Chao Chen, Zhixin Ma, Yongqi Li, Yupeng Hu, Yinwei Wei, Wenjie Li, Liqiang Nie
https://arxiv.org/abs/2510.12603

Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space
Multimodal reasoning aims to enhance the capabilities of MLLMs by incorporating intermediate reasoning steps before reaching the final answer. It has evolved from text-only reasoning to the integration of visual information, enabling the thought process to be conveyed through both images and text. Despite its effectiveness, current multimodal reasoning methods depend on explicit reasoning steps that require labor-intensive vision-text annotations and inherently introduce significant inference l…

Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
Ruben Belo, Claudia Soares, Marta Guimaraes
https://arxiv.org/abs/2510.12672

Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
Large Language Models are susceptible to jailbreak attacks that bypass built-in safety guardrails (e.g., by tricking the model with adversarial prompts). We propose Concept Alignment and Concept Manipulation \textbf{CALM}, an inference-time method that suppresses harmful concepts by modifying latent representations of the last layer of the model, without retraining. Leveraging \gls*{cw} technique from Computer Vision combined with orthogonal projection, CALM removes unwanted latent directions a…