Tootfinder

The number of arguments is unimportant unless some of them are correct.
-- Ralph Hartley

The Unwinnable Arms Race of AI Image Detection
Till Aczel, Lorenzo Vettor, Andreas Plesner, Roger Wattenhofer
https://arxiv.org/abs/2509.21135 https://arxi…

The Unwinnable Arms Race of AI Image Detection
The rapid progress of image generative AI has blurred the boundary between synthetic and real images, fueling an arms race between generators and discriminators. This paper investigates the conditions under which discriminators are most disadvantaged in this competition. We analyze two key factors: data dimensionality and data complexity. While increased dimensionality often strengthens the discriminators ability to detect subtle inconsistencies, complexity introduces a more nuanced effect. Usi…

Predicting the Order of Upcoming Tokens Improves Language Modeling
Zayd M. K. Zuhri, Erland Hilman Fuadi, Alham Fikri Aji
https://arxiv.org/abs/2508.19228 https://

Predicting the Order of Upcoming Tokens Improves Language Modeling
Multi-Token Prediction (MTP) has been proposed as an auxiliary objective to improve next-token prediction (NTP) in language model training but shows inconsistent improvements, underperforming in standard NLP benchmarks. We argue that MTP's exact future token prediction is too difficult as an auxiliary loss. Instead, we propose Token Order Prediction (TOP), which trains models to order upcoming tokens by their proximity using a learning-to-rank loss. TOP requires only a single additional unembed…

from my link log —
Renewable energy unambiguously reduces UK wholesale power prices.
https://www.carboncommentary.com/blog/2025/8/18/renewables-do-unambiguously-reduce-wholesale-power-prices
saved 2025-08-20

Renewables do unambiguously reduce wholesale power prices | Carbon Commentary
We still hear assertions that adding renewables to the grid has increased the UK’s electricity costs. I looked at two sources of data and plotted one against the other to test whether there’s any truth in this. 1,      The ‘next day’ electricity price fo

A century ago, the government hired unemployed young men to build America's forests, trails, and parks. Photos show FDR's 'tree army.' - Business Insider
https://www.businessinsider.com/civilian-conservation-corps-ccc-fdr-parks-forests-photos

Raiders report: Ashton Jeanty unimpressed by viral run, Madden rating https://www.reviewjournal.com/sports/raiders/raiders-report-ashton-jeanty-unimpressed-by-viral-run-madden-rating-3426638/

In ant colonies infected by a tapeworm parasite, uninfected workers show signs of stress-related gene activity, in part because they're caring for infected nestmates (Photo by Matt Hamer, via AntWiki)
Parasite prevalence in a social host has colony-wide impacts on transcriptional activity and survival
https://doi.org/10.1093/ev…

The Atlanta Fed estimates delays faced by claimants in COBOL-based unemployment insurance systems cut US GDP by at least $40B between March 13 and year-end 2020 (Bryce Elder/Financial Times)
https://www.ft.com/content/7610a755-27e3-4792-a337-27d9b301000b

NYC Tonight!
I’ll be reading at Unnameable Books in Brooklyn.
Come for Filip Marinovich, The Ghost Shepherds, Eddie Berrigan and our host, Bathtub Nebula — stay for me!
Thursday September 18 7:30-9:30 Unnameable Books 615 Vanderbilt Ave. Brooklyn
#poetry
—

humancompatible.train: Implementing Optimization Algorithms for Stochastically-Constrained Stochastic Optimization Problems
Andrii Kliachkin, Jana Lep\v{s}ov\'a, Gilles Bareilles, Jakub Mare\v{c}ek
https://arxiv.org/abs/2509.21254

humancompatible.train: Implementing Optimization Algorithms for Stochastically-Constrained Stochastic Optimization Problems
There has been a considerable interest in constrained training of deep neural networks (DNNs) recently for applications such as fairness and safety. Several toolkits have been proposed for this task, yet there is still no industry standard. We present humancompatible.train (https://github.com/humancompatible/train), an easily-extendable PyTorch-based Python package for training DNNs with stochastic constraints. We implement multiple previously unimplemented algorithms for stochastically constra…