Tootfinder

In some UK woodlands, every young tree has died. What’s going wrong? https://www.theguardian.com/environment/2025/jul/10/britain-ancient-woodlands-failing-regenerate-forests-climate-drought-heat-dis…

Det här är en studie där man mätt omega-6 nivåer i blodet och sen tittat på korrelationer. Men det här är även vad interventionsstudier och tidigare bättre observationsstudier också har visat.
Seed oils: Omega-6 fatty acids may lower diabetes, heart disease risk https://www.

A Generative Imputation Method for Multimodal Alzheimer's Disease Diagnosis
Reihaneh Hassanzadeh, Anees Abrol, Hamid Reza Hassanzadeh, Vince D. Calhoun
https://arxiv.org/abs/2508.09271

A Generative Imputation Method for Multimodal Alzheimer's Disease Diagnosis
Multimodal data analysis can lead to more accurate diagnoses of brain disorders due to the complementary information that each modality adds. However, a major challenge of using multimodal datasets in the neuroimaging field is incomplete data, where some of the modalities are missing for certain subjects. Hence, effective strategies are needed for completing the data. Traditional methods, such as subsampling or zero-filling, may reduce the accuracy of predictions or introduce unintended biases.…

Behind the Blog: Exercises in OSINT and Storage Pains https://www.404media.co/behind-the-blog-exercises-in-osint-and-storage-pains/

Condition number for finite element discretisation of nonlocal PDE systems with applications to biology
Olusegun E. Adebayo, Raluca Eftimie, Dumitru Trucu
https://arxiv.org/abs/2508.09781

Condition number for finite element discretisation of nonlocal PDE systems with applications to biology
In this work, we investigate the condition number for a system of coupled non-local reaction-diffusion-advection equations developed in the context of modelling normal and abnormal wound healing. Following a finite element discretisation of the coupled non-local system, we establish bounds for this condition number. We further discuss how model parameter choices affect the conditioning of the system. Finally, we discuss how the step size of the chosen time-stepping scheme and the spatial gr…

Der ConniMeme Instagram-Account, den der Carlsen-Verlag hat löschen lassen, ist wieder da. 🥳
Keine Ahnung für wie lange, aber es reicht hoffentlich um alles zu crawlen 🙂
https://www.instagram.com/connimeme/

Analytic model for neutral penetration and plasma fueling
George J. Wilkie
https://arxiv.org/abs/2506.10906 https://arxiv.org/pdf/250…

Analytic model for neutral penetration and plasma fueling
Neutral atoms recycled from wall interaction interact with confined plasma, thereby refueling it, most strongly in the region closest to the wall. This occurs near the X-point in diverted configurations, or else near the wall itself in limited configurations. A progression of analytic models are developed for neutral density in the vicinity of a planar or linear source in an ionizing domain. First-principles neutral transport simulations with DEGAS2 are used throughout to test the validity and …

I share a lot of @…’s feelings here. (He doesn’t mention training data sourcing, but that’s on my mind too.)
There’s a layer of ethics and social impact that hangs, as Tim put it, as a stench over all the thinking about whether and how the tech is useful. I am actually willing to discuss the utility questions, but they are secondary.
https://cosocial.ca/@timbray/114807889514002925

Right now there's an ACM design technology conference happening with hundreds of people flying in to a conference facility on a remote island to discuss the conference theme of how to make their work more sustainable. In-person participation only.

Generalizing Verifiable Instruction Following
Valentina Pyatkin, Saumya Malik, Victoria Graf, Hamish Ivison, Shengyi Huang, Pradeep Dasigi, Nathan Lambert, Hannaneh Hajishirzi
https://arxiv.org/abs/2507.02833

Generalizing Verifiable Instruction Following
A crucial factor for successful human and AI interaction is the ability of language models or chatbots to follow human instructions precisely. A common feature of instructions are output constraints like ``only answer with yes or no" or ``mention the word `abrakadabra' at least 3 times" that the user adds to craft a more useful answer. Even today's strongest models struggle with fulfilling such constraints. We find that most models strongly overfit on a small set of verifiable constraints from …