Tootfinder

lmao

Sliding-Window Signatures for Time Series: Application to Electricity Demand Forecasting
Nina Drobac (LPSM), Margaux Br\'eg\`ere (LPSM), Joseph de Vilmarest (LPSM), Olivier Wintenberger (LPSM)
https://arxiv.org/abs/2510.12337

Sliding-Window Signatures for Time Series: Application to Electricity Demand Forecasting
Nonlinear and delayed effects of covariates often render time series forecasting challenging. To this end, we propose a novel forecasting framework based on ridge regression with signature features calculated on sliding windows. These features capture complex temporal dynamics without relying on learned or hand-crafted representations. Focusing on the discrete-time setting, we establish theoretical guarantees, namely universality of approximation and stationarity of signatures. We introduce an …

Finished "Lobizona" by Romina Garber. I have extremely mixed feelings about this book. It's a powerful depiction of the fear of living as an undocumented child/teen and it has interesting things to say about rejection, belonging, and the choice between seeking to be recognized for who you are and wanting you blend in enough to be accepted as normal. However, it's also an explicit homage to Harry Potter, and while it doesn't include antisemitic tropes or glorify slavery or even have any anti-trans sentiments I can detect, to me the magical school setup felt forced and I thought it would have been a better book had it not tried to fit that mould. Also, it would have been a super interesting situation to explore trans issues, and while it's definitely fine for it not to do that, the author's praise of Rowling's work has me wondering...
There's a sequel that I think could in theory be amazing, but given the execution of the first book, I think I'll wait a bit before checking it out. By putting her main character in opposition to both ICE in the human world and the magical authorities in the other world, Garber explicitly sets the stage for a revolution standing between her protagonist and any kind of lasting peace. But I'm not confident she's capable of writing that story without relying on some kind of supernatural deus ex machina, which would be disappointing to me, since "a better world if only possible through divine intervention" is an inherently regressive message.
Overall, #OwnVoices fantasy centering an undocumented immigrant is an excellent thing, and I've certainly got a lot of privilege that surely influences my criticism. However, #OwnVoices stuff has a range of levels of craft and political stances, and it can be excellent for some reasons and mediocre for others.
On that point, if anyone reading this has suggestions for fiction books grappling with borders and the carceral state, Is be happy to hear them.
#AmReading

Reionization in HESTIA: Studying reionization in the LG through zoom simulations
David Attard, Luke Conaboy, Noam Libeskind, Sergey Pillipenko, Keri Dixon, Ilian T. Iliev
https://arxiv.org/abs/2509.10133

Reionization in HESTIA: Studying reionization in the LG through zoom simulations
While cosmic reionization has been broadly constrained by global observables, the interplay between internal sources (Milky Way, M31, and their satellites) and external ionization fronts remains poorly understood in a realistic Local Group (LG) context. To address this issue, we perform radiative transfer post-processing on the original HESTIA LG constrained simulation. We calibrate our source models using a uniform 1024^3 particle, dark-matter only, HESTIA simulation coupled with a subgrid col…

Mic Drop or Data Flop? Evaluating the Fitness for Purpose of AI Voice Interviewers for Data Collection within Quantitative & Qualitative Research Contexts
Shreyas Tirumala, Nishant Jain, Danny D. Leybzon, Trent D. Buskirk
https://arxiv.org/abs/2509.01814

Mic Drop or Data Flop? Evaluating the Fitness for Purpose of AI Voice Interviewers for Data Collection within Quantitative & Qualitative Research Contexts
Transformer-based Large Language Models (LLMs) have paved the way for "AI interviewers" that can administer voice-based surveys with respondents in real-time. This position paper reviews emerging evidence to understand when such AI interviewing systems are fit for purpose for collecting data within quantitative and qualitative research contexts. We evaluate the capabilities of AI interviewers as well as current Interactive Voice Response (IVR) systems across two dimensions: input/output perform…

If you have a green card and get a single pot conviction, you have to either naturalize or never travel internationally.
Here's the relevant section from a legal practice advisory on this issue.
https://bsky.app/profile/reichlinmelnick.bsky.social/post/3lv47…

Aaron Reichlin-Melnick (@reichlinmelnick.bsky.social)
Yep, but he travelled abroad, which is what got him. That made him "inadmissible," not "deportable." So if you have a green card and get a single pot conviction, you have to either naturalize or never travel internationally. Here's the relevant section from a legal practice advisory on this issue. [contains quote post or other embedded content]

Scientists create "solar" cells capable of using ambient indoor light.
"Your household gadgets could soon be battery-free..."
https://www.livescience.com/technology/your-gadgets-could-soo…

Your household gadgets could soon be battery-free — scientists create tiny solar cells that can be powered by indoor light
Researchers said the breakthrough "paves the way for electronics powered by the ambient light already present in our lives."

Evaluating LLMs for Historical Document OCR: A Methodological Framework for Digital Humanities
Maria Levchenko
https://arxiv.org/abs/2510.06743 https://arx…

Evaluating LLMs for Historical Document OCR: A Methodological Framework for Digital Humanities
Digital humanities scholars increasingly use Large Language Models for historical document digitization, yet lack appropriate evaluation frameworks for LLM-based OCR. Traditional metrics fail to capture temporal biases and period-specific errors crucial for historical corpus creation. We present an evaluation methodology for LLM-based historical OCR, addressing contamination risks and systematic biases in diplomatic transcription. Using 18th-century Russian Civil font texts, we introduce novel …

Share Your Attention: Transformer Weight Sharing via Matrix-based Dictionary Learning
Magauiya Zhussip, Dmitriy Shopkhoev, Ammar Ali, Stamatios Lefkimmiatis
https://arxiv.org/abs/2508.04581

Share Your Attention: Transformer Weight Sharing via Matrix-based Dictionary Learning
Large language models (LLMs) have revolutionized AI applications, yet their high computational and memory demands hinder their widespread deployment. Existing compression techniques focus on intra-block optimizations (e.g. low-rank approximation, attention head pruning), while the repetitive layered structure of transformers implies significant inter-block redundancy - a dimension largely unexplored beyond key-value (KV) caching. Inspired by dictionary learning in CNNs, we propose a framework f…

When the Spanish colonized the region in the 17th century,
they didn't introduce horses to Indigenous people, as long thought.
Instead, horses were present in the Southwest long before Europeans,
and were traded by Indigenous people who formed close relationships with them.
Horses lived in North America for millions of years
but went extinct at the end of the last ice age, about 11,000 years ago.
When Europeans reintroduced horses to what is now the ea…

Indigenous people of the American West used 'sacred' horses a half-century earlier than previously thought
Indigenous oral histories and archaeological evidence are rewriting the story of how horses came to the American West.