Tootfinder

The paradox of Nazism being both the most capitalist thing possible, defining itself as explicitly anticommunist, while masquerading as "socialism" is difficult to resolve until you understand one thing: fascism pivots around antisemitism (and its extension, conspiratorial thinking).
I'm, of course, not talking about the redefinition of "antisemitism" into meaning any criticism of Israel but rather an anti-Jewish conspiracy narrative rooted in Roman Christianity.
This is critical to understand as MAGA fascism pivots between capitalist and pseudo-anticapitalist with Trump in the middle. Hitler did the same thing. Strasserism helped the Nazis gain power by pulling in the Left. In the Nazis case they killed the Stasserists pretty quickly. Now, I think we're seeing an attempt to make the opposite pivot happen in MAGA. But it's all the same thing.
Fascism can infinitely fail to address the needs of the people while dismissing it's own responsibility for creating the problem by maintaining a permanent enemy.
This is why it's important to understand antisemitism and how to fight it. It's especially important now because the apparatus of violence in Israel is itself a tool of global fascism, and we finally have an opportunity to dismantle the whole thing. But we have to be aware of how fascists can pivot around to block this.
I've been reading Safety Through Solidarity, and I think it's especially relevant at this time.
https://www.akpress.org/safety-through-solidarity.html

Falls jemand hier im Bereich Software Entwicklung / Prototyping unterwegs ist, in Berlin wohnt (oder wohnen wollen würde) und nen Job sucht, wir haben gerade openings. (Nicht nur die explizit ausgeschriebene Junior Stelle, auch Seniors könnten relevant sein) https://artcom.de/?job_description=jun…

Physio Network
Continues to bring you the biggest names in physio, tackling the most clinically relevant topics...
Great Australian Pods Podcast Directory: https://www.greataustralianpods.com/physio-explained/

CDE: Concept-Driven Exploration for Reinforcement Learning
Le Mao, Andrew H. Liu, Renos Zabounidis, Zachary Kingston, Joseph Campbell
https://arxiv.org/abs/2510.08851 https://…

CDE: Concept-Driven Exploration for Reinforcement Learning
Intelligent exploration remains a critical challenge in reinforcement learning (RL), especially in visual control tasks. Unlike low-dimensional state-based RL, visual RL must extract task-relevant structure from raw pixels, making exploration inefficient. We propose Concept-Driven Exploration (CDE), which leverages a pre-trained vision-language model (VLM) to generate object-centric visual concepts from textual task descriptions as weak, potentially noisy supervisory signals. Rather than direct…

Toward Mechanistic Explanation of Deductive Reasoning in Language Models
Davide Maltoni, Matteo Ferrara
https://arxiv.org/abs/2510.09340 https://arxiv.org/…

Toward Mechanistic Explanation of Deductive Reasoning in Language Models
Recent large language models have demonstrated relevant capabilities in solving problems that require logical reasoning; however, the corresponding internal mechanisms remain largely unexplored. In this paper, we show that a small language model can solve a deductive reasoning task by learning the underlying rules (rather than operating as a statistical learner). A low-level explanation of its internal representations and computational circuits is then provided. Our findings reveal that inducti…

Gods, let this be the thing that finally ends his ordeal and sets him free.
https://innocenceproject.org/news/breaking-explosive-new-evidence-revealed-yesterday-in-n…

BREAKING: Explosive New Evidence Revealed in NBC “Dateline” Podcast Related to Pending Claim in Robert Roberson Case - Innocence Project
Filing with Texas Court of Criminal Appeals Says Newly Available Evidence Proves Judicial Misconduct, Requires New Trial for Robert Roberson

Relevant Deformations, Brane Brick Models and Triality
Mario Carcamo, Sebastian Franco, Dongwook Ghim, Georgios P. Goulas, Rak-Kyeong Seong
https://arxiv.org/abs/2510.05517 http…

Relevant Deformations, Brane Brick Models and Triality
We extend the study of relevant deformations connecting 2d (0,2) gauge theories on D1-branes probing toric Calabi-Yau 4-folds beyond pure mass deformations. The underlying geometry provides powerful insights when field-theoretic tools are still lacking. We observe that the volume of the Sasaki-Einstein base of the Calabi-Yau 4-fold grows towards the IR, signaling the relevance of deformations. We exploit the map between gauge theory fields and GLSM fields to compute scaling dimensions directly …

Imagine ChatGPT but instead of predicting text it just linked you to the to 3 documents most-influential on the probabilities that would have been used to predict that text.
Could even generate some info about which parts of each would have been combined how.
There would still be issues with how training data is sourced and filtered, but these could be solved by crawling normally respecting robots.txt and by paying filterers a fair wage with a more relaxed work schedule and mental health support.
The energy issues are mainly about wild future investment and wasteful query spam, not optimized present-day per-query usage.
Is this "just search?"
Yes, but it would have some advantages for a lot of use cases, mainly in synthesizing results across multiple documents and in leveraging a language model more fully to find relevant stuff.
When we talk about the harms of current corporate LLMs, the opportunity cost of NOT building things like this is part of that.
The equivalent for art would have been so amazing too! "Here are some artists that can do what you want, with examples pulled from their portfolios."
It would be a really cool coding assistant that I'd actually encourage my students to use (with some guidelines).
#AI #GenAI #LLMs

SliceFine: The Universal Winning-Slice Hypothesis for Pretrained Networks
Md Kowsher, Ali O. Polat, Ehsan Mohammady Ardehaly, Mehrdad Salehi, Zia Ghiasi, Prasanth Murali, Chen Chen
https://arxiv.org/abs/2510.08513

SliceFine: The Universal Winning-Slice Hypothesis for Pretrained Networks
This paper presents a theoretical framework explaining why fine tuning small, randomly selected subnetworks (slices) within pre trained models can be sufficient for downstream adaptation. We prove that pretrained networks exhibit a universal winning slice property arising from two phenomena: (1) spectral balance the eigenspectra of different weight matrix slices are remarkably similar; and (2) high task energy their backbone representations retain rich, task relevant features. This leads to the…

Artificial Intelligence Virtual Cells: From Measurements to Decisions across Modality, Scale, Dynamics, and Evaluation
Chengpeng Hu, Calvin Yu-Chian Chen
https://arxiv.org/abs/2510.12498

Artificial Intelligence Virtual Cells: From Measurements to Decisions across Modality, Scale, Dynamics, and Evaluation
Artificial Intelligence Virtual Cells (AIVCs) aim to learn executable, decision-relevant models of cell state from multimodal, multiscale measurements. Recent studies have introduced single-cell and spatial foundation models, improved cross-modality alignment, scaled perturbation atlases, and explored pathway-level readouts. Nevertheless, although held-out validation is standard practice, evaluations remain predominantly within single datasets and settings; evidence indicates that transport acr…