Tootfinder

Originally, I argued to skip the «book of abstract» for SIG Writing 26. And maybe, creating one will acually be the best part in organizing it as I hopefully can do it with a fellow document engineer with a proper XML–XSLT–XSLT-FO–HTML–ePub pipeline! Maybe even LaTeX, who knows
Or the second best thing after printing stickers :)
#WritingResearch

Efficient Mixed-Precision Large Language Model Inference with TurboMind
Li Zhang, Youhe Jiang, Guoliang He, Xin Chen, Han Lv, Qian Yao, Fangcheng Fu, Kai Chen
https://arxiv.org/abs/2508.15601

Efficient Mixed-Precision Large Language Model Inference with TurboMind
Mixed-precision inference techniques reduce the memory and computational demands of Large Language Models (LLMs) by applying hybrid precision formats to model weights, activations, and KV caches. This work introduces mixed-precision LLM inference techniques that encompass (i) systematic memory and compute optimization across hierarchical storage and tensor core architectures, and (ii) comprehensive end-to-end mixed-precision optimization across diverse precision formats and hardware configurati…

Hijacking Living Cells with Surface Engineering for the Internet of Bio-Nano Things
Ekin Ince, Murat Kuscu
https://arxiv.org/abs/2509.17227 https://arxiv.o…

Hijacking Living Cells with Surface Engineering for the Internet of Bio-Nano Things
The Internet of Bio-Nano Things (IoBNT) promises to revolutionize healthcare by interfacing the cyber domain with the living systems at unprecedented resolution. Realizing this vision hinges on the development of Bio-Nano Things (BNTs), i.e., functional nodes capable of sensing, actuation, and communications within biological environments. Existing BNT architectures, e.g., nanomaterial-based, biosynthetic, and passive molecular agents, face significant limitations, including toxicity, lack of a…

Dumping Google’s enshittified search for Kagi - Ars Technica https://arstechnica.com/gadgets/2025/08/enough-is-enough-i-dumped-googles-worsening-search-for-kagi/

Enough is enough—I dumped Google’s worsening search for Kagi
Enough is enough: I’ve jettisoned Google in favor of a search engine that doesn’t treat me like a product.

Just to hammer this home a bit more, I knew a guy who joined the army. I met him after he went AWOL. When he went into the recruiter's office, they asked him if he had any open warrants out for his arrest because they couldn't recruit anyone with a criminal record. He said he did, then they said, "oh, actually, we can help you with all that, don't get caught before we ship you out."
He was just trying to keep himself out of jail. That's not supposed to happen, but it does. IIRC it was on a drug charge, which, also, they're not supposed to take anyone who tests positive for weed... but they also just tell you how to prepare for a drug test.
Another friend joined because she wanted to be part Army Corps of Engineers. A good chunk of the folks I went to school with joined the military after graduation because the other choices were working at the saw mill or working at the canary. If you join the military, you get to go to college. (Or you get to stay out of jail... as long as you don't go AWOL.)

Constrained Sensing and Reliable State Estimation with Shallow Recurrent Decoders on a TRIGA Mark II Reactor
Stefano Riva, Carolina Introini, Jos\`e Nathan Kutz, Antonio Cammi
https://arxiv.org/abs/2510.12368

Constrained Sensing and Reliable State Estimation with Shallow Recurrent Decoders on a TRIGA Mark II Reactor
Shallow Recurrent Decoder networks are a novel data-driven methodology able to provide accurate state estimation in engineering systems, such as nuclear reactors. This deep learning architecture is a robust technique designed to map the temporal trajectories of a few sparse measures to the full state space, including unobservable fields, which is agnostic to sensor positions and able to handle noisy data through an ensemble strategy, leveraging the short training times and without the need for …

Cool little adapted tale on AI, by Alison Gopnik:
https://simons.berkeley.edu/news/stone-soup-ai

Stone Soup AI
For some time, I’ve argued that a common conception of AI is misguided. This is the idea that AI systems like large language and vision models are individual intelligent agents, analogous to human agents. Instead, I’ve argued that these models are “cultural technologies” like writing, print, pictures, libraries, internet search engines, and Wikipedia.

#OTD: 13 years ago, I visited the Mercedes-Benz Museum in Stuttgart.
It is one of Germany's temples to “das Auto”. Architectural gems, this one is by UN Studio.

PUUMA (Placental patch and whole-Uterus dual-branch U-Mamba-based Architecture): Functional MRI Prediction of Gestational Age at Birth and Preterm Risk
Diego Fajardo-Rojas, Levente Baljer, Jordina Aviles Verdera, Megan Hall, Daniel Cromb, Mary A. Rutherford, Lisa Story, Emma C. Robinson, Jana Hutter
https://arxiv.org/abs/2509.07042

PUUMA (Placental patch and whole-Uterus dual-branch U-Mamba-based Architecture): Functional MRI Prediction of Gestational Age at Birth and Preterm Risk
Preterm birth is a major cause of mortality and lifelong morbidity in childhood. Its complex and multifactorial origins limit the effectiveness of current clinical predictors and impede optimal care. In this study, a dual-branch deep learning architecture (PUUMA) was developed to predict gestational age (GA) at birth using T2* fetal MRI data from 295 pregnancies, encompassing a heterogeneous and imbalanced population. The model integrates both global whole-uterus and local placental features. I…