Tootfinder

An excellent summary here via @… about how the structures of our civil society have failed to stop ICE from becoming Trump’s Brownshirts.
The one thing the piece omits: someone •is• stopping ICE. It’s the citizens filling the streets, honking and shouting and filming and generally harassing ICE, doing the work our government has failed to do. If it were not for that response being so widespread, sustained, and forceful, we’d be in far worse place right now. https://mastodon.social/@AnnaAnthro/115616389777857234

Flame Skimmer Dragonfly hanging onto a branch with its mandibles. LA Arboretum, Arcadia, California, USA. October 2025. OM System OM-1 M.Zuiko 300mm F4. #laarboretum #dragonfly #flameskimmer

Moody Urbanity - Stand 🧍‍♂️
情绪化城市 - 立 🧍‍♂️
📷 Nikon FE
🎞️ Ilford HP5 Plus 400, expired 1993
#filmphotography #Photography #blackandwhite

This tree ornament sure is hard to photograph

Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformers
Gangwei Xu, Haotong Lin, Hongcheng Luo, Xianqi Wang, Jingfeng Yao, Lianghui Zhu, Yuechuan Pu, Cheng Chi, Haiyang Sun, Bing Wang, Guang Chen, Hangjun Ye, Sida Peng, Xin Yang
https://arxiv.org/abs/2510.07316

Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformers
This paper presents Pixel-Perfect Depth, a monocular depth estimation model based on pixel-space diffusion generation that produces high-quality, flying-pixel-free point clouds from estimated depth maps. Current generative depth estimation models fine-tune Stable Diffusion and achieve impressive performance. However, they require a VAE to compress depth maps into latent space, which inevitably introduces \textit{flying pixels} at edges and details. Our model addresses this challenge by directly…

Ultrathin bismuth-yttrium iron garnet films with tunable and compensated magnetic anisotropy
Hanchen Wang, William Legrand, Davit Petrosyan, Min-Gu Kang, Emir Karad\v{z}a, Hiroki Matsumoto, Richard Schlitz, Michaela Lammel, Myriam H. Aguirre, Pietro Gambardella
https://arxiv.org/abs/2510.07465

Ultrathin bismuth-yttrium iron garnet films with tunable and compensated magnetic anisotropy
We report on the epitaxial growth of nm-thick films of bismuth-substituted yttrium iron garnet (BiYIG) by high-temperature off-axis radio-frequency magnetron sputtering. We demonstrate accurate control of the magnetic properties by tuning of the sputtering parameters and epitaxial strain on various (111)-oriented garnet substrates. BiYIG films with up to -0.80\% lattice mismatch with the substrate remain fully strained up to 60~nm-thick, maintaining a high crystalline quality. Transmission elec…

Short day today as the worker had to leave early for some sort of appointment.
Still, the hanging rails are up inside the wardrobe, the racks are up in the closet to support the adjustable shelves in there, the fixed shelves at the side of the cupboard are up.
I thought I smelled paint for a while before he left but turns out that was "wood filler", apparently filling the holes in the natural wood?
Not sure what this frame will be. Is it the edges for a panel door? Is it the start of the windowed doors for the top cupboards? Dunno.

R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?
Yi Lu, Jianing Wang, Linsen Guo, Wei He, Hongyin Tang, Tao Gui, Xuanjing Huang, Xuezhi Cao, Wei Wang, Xunliang Cai
https://arxiv.org/abs/2510.08189

R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?
Recent trends in test-time scaling for reasoning models (e.g., OpenAI o1, DeepSeek-R1) have led to remarkable improvements through long Chain-of-Thought (CoT). However, existing benchmarks mainly focus on immediate, single-horizon tasks, failing to adequately evaluate models' ability to understand and respond to complex, long-horizon scenarios. To address this incomplete evaluation of Large Reasoning Models (LRMs), we propose R-HORIZON, a method designed to stimulate long-horizon reasoning beha…

Decoupled Multimodal Fusion for User Interest Modeling in Click-Through Rate Prediction
Alin Fan, Hanqing Li, Sihan Lu, Jingsong Yuan, Jiandong Zhang
https://arxiv.org/abs/2510.11066

Decoupled Multimodal Fusion for User Interest Modeling in Click-Through Rate Prediction
Modern industrial recommendation systems improve recommendation performance by integrating multimodal representations from pre-trained models into ID-based Click-Through Rate (CTR) prediction frameworks. However, existing approaches typically adopt modality-centric modeling strategies that process ID-based and multimodal embeddings independently, failing to capture fine-grained interactions between content semantics and behavioral signals. In this paper, we propose Decoupled Multimodal Fusion (…

Flame Skimmer dragonfly resting on a green branch. LA Arboretum, Arcadia, California, USA. #laarboretum #dragonfly #flameskimmer