Tootfinder

Nvidia announces its Earth-2 Medium Range weather model, built on its Atlas architecture, claiming it outperforms Google DeepMind's GenCast in 70 variables (Tim De Chant/TechCrunch)
https://techcrunch.com/2026/01/26/nvidias-new-a…

Great post by @… on the PlayStation 2 and its design architecture: https://machaddr.com/blog/articles/2026-01-20-playstation-2-internals

The new Zygote issue 39 is featuring stunning art by Kate MccGwire[1] and articles about:
Forward Motion: Robert J. Full’s Career of Discovery
Science of Seeing: Plant It and They Will Come
Cutting Without Cutting: Lessons from the Sawfly
Interview with Arindam Phani and Seonghwan (Sam) Kim
Building with Biology
Turning Bio-Inspired Ideas Into Patents
Biomimicry in Architecture

"ps. m88k has been the most worthwhile example, since some insane parts
of that architecture permits Miod to find a MI bug weekly."
I just live that quote.
https://marc.info/?l=openbsd-misc&m=109160932504578&w=2

Boring Technology, Boring Architecture
#systemarchitecture

We’re at the @… today with a booth of our own. Will also give a talk on Glorious Data Models later today.
https://

Series D, Episode 03 - Traitor
DAYNA: I didn't know Helots were originally from Earth.
VILA: Everyone came from Earth originally. That's a well-known fact.
SOOLIN: It's a well-known opinion, actually.
TARRANT: Most well known facts are.
https://blake.torpidity.net/m/403/96<…

from my link log —
Software architecture is overrated, clear and simple design is underrated.
https://blog.pragmaticengineer.com/software-architecture-is-overrated/
saved 2019-09-19

i gave my son-in-law a battery daddy (fully populated), a scrub daddy set, and a gift certificate to fat daddy
#BatteryDaddy #ScrubDaddy #FatDaddy

Moody Urbanity - Old & New II 🔆
情绪化城市 - 新与旧 II 🔆
📷 Nikon FE
🎞️ Ilford HP5 Plus 400, expired 1993
#filmphotography #Photography #blackandwhite

🏗️ Problem 4: Architecture decay. Classes become dumping grounds for random data: $user->cached_data, $user->temp_flag, $user->random_stuff. Code becomes unmaintainable.
✅ Modern PHP solution: Declare all properties explicitly with types. Use public string $name; public int $age; public bool $isActive; in your class definitions.

Since the so-called ceasefire came into effect in Gaza on October 10,
Israel has been consolidating its control of over 50% of Gaza and
—according to new research by Forensic Architecture
—physically altering the geography of the land.
Through a combination of the construction of military infrastructure alongside the destruction of existing buildings,
Israel appears to be laying the groundwork to establish a permanent presence in the majority of the Gaza Strip.

More shots from today. #Chicago doesn't have a bad side ;)
#photography #Architecture

EUROPEAN LEADERS discuss new SECURITY ARCHITECTURE #shorts: https://benborges.xyz/2026/01/22/european-leaders-discuss-new-security.html

Ich freue mich, mal wieder beim Software Architecture Summit in München dabei zu sein - ich mache einen Workshop zu "Data-Oriented Programming”, mithin zu High-Level-Domänenmodellierung.
https://software-architecture-summit.de/modeling-und-design/da…

Two favourite sessions in a row with my MSc students at the Bartlett School of Architecture discussing buildings and social relationships 🫶🏼
Short 🧵
#ArchitecturalResearch #sociology #architecture

Neuroplasticity: Limits and the architecture of change https://medium.com/@krigerbruce/neuroplasticity-limits-and-the-architecture-of-change-be5346cae7f8 Case study of sensory substitution: Visual cortex activation in congenital blindne…

@chadfowler.com's second post in his new Phoenix Architecture blog, called The Death and Rebirth of Programming.
Lots of quotable quotes, so hard to choose. "Soft skills" more important than ever, and this is all be a shock to people who self-identify as programmers.
https://aicoding…

Moody Urbanity - Past 📴
情绪化城市 - 过去 📴
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Ilford HP5, expired 1993
#filmphotography #Photography #blackandwhite

So ... which simulation of the large scale structure of the #Universe is this?
(Haven't found a way to hide the solution but not the image, to make it a quiz.)
This is from the press release "Observing synapses in action" - https://www.charite.de/en/service/press_reports/artikel/detail/observing_synapses_in_action - about the paper "Dynamic nanoscale architecture of synaptic vesicle fusion in mouse hippocampal neurons", https://www.nature.com/articles/s41467-025-67291-6 ;-)

🧠 Human-like Working Memory from Artificial Intrinsic Plasticity Neurons
#ai #llm

Series B, Episode 04 - Horizon
ZEN: Planet visual is now available. [Horizon appears on main screen]
JENNA: Still holding course, Standard by Two.
BLAKE: Freighter's speed?
ZEN: Time Distort Six.
BLAKE: Freighter's planetfall?
https://blake.torpidity.net/m/204/69

The #CleanArchitecture I Wish Someone Had Explained to Me
https://medium.com/@rafael-22/the-clean-architecture-i-wish-someone-had-expl…

🤖 Designed for AI coding agents like #GeminiCLI, #ClaudeCode, and others with system prompt integration and Agent Skills support
📋 Three-subcommand architecture: info for discovery, grep for searching, call for execution - structured error messages with recovery suggestions
🔧 …

Multi-port programmable silicon photonics using low-loss phase change material Sb$_2$Se$_3$
Thomas W. Radford, Idris A Ajia, Latif Rozaqi, Priya Deoli, Xingzhao Yan, Mehdi Banakar, David J Thomson, Ioannis Zeimpekis, Alberto Politi, Otto L. Muskens
https://arxiv.org/abs/2511.18205 https://arxiv.org/pdf/2511.18205 https://arxiv.org/html/2511.18205
arXiv:2511.18205v1 Announce Type: new
Abstract: Reconfigurable photonic devices are rapidly emerging as a cornerstone of next generation optical technologies, with wide ranging applications in quantum simulation, neuromorphic computing, and large-scale photonic processors. A central challenge in this field is identifying an optimal platform to enable compact, efficient, and scalable reconfigurability. Optical phase-change materials (PCMs) offer a compelling solution by enabling non-volatile, reversible tuning of optical properties, compatible with a wide range of device platforms and current CMOS technologies. In particular, antimony tri-selenide ($\text{Sb}_{2}\text{Se}_{3}$) stands out for its ultra low-loss characteristics at telecommunication wavelengths and its reversible switching. In this work, we present an experimental platform capable of encoding multi-port operations onto the transmission matrix of a compact multimode interferometer architecture on standard 220~nm silicon photonics using \textit{in-silico} designed digital patterns. The multi-port devices are clad with a thin film of $\text{Sb}_{2}\text{Se}_{3}$, which can be optically addressed using direct laser writing to provide local perturbations to the refractive index. A range of multi-port geometries from 2$\times$2 up to 5$\times$5 couplers are demonstrated, achieving simultaneous control of up to 25 matrix elements with programming accuracy of 90% relative to simulated patterns. Patterned devices remain stable with consistent optical performance across the C-band wavelengths. Our work establishes a pathway towards the development of large scale PCM-based reconfigurable multi-port devices which will allow implementing matrix operations on three orders of magnitude smaller areas than interferometer meshes.
toXiv_bot_toot

Check out my latest article - "The Psychological Architecture of Hope in a Warming World"
https://www.brichapman.com/p/the-psychological-architecture-of?utm_source=bri-mastodon&utm_medium=social

Alle acht bisherigen Vorträge der Reihe „Open Divide – Critical Studies on Open Access“ (weitere werden folgen) sind in der Mediathek der TIB verfügbar:
#OpenAccess

Flowing architecture by Frank Gehry #RIP #architecture

Look what an old #macOS SDK in the “Command Line Tools" tree causes on Apple Silicon:
$ clang -o ghba geh.c
In file included from geh.c:14:
In file included from /Library/Developer/CommandLineTools/SDKs/MacOSX10.14.sdk/usr/include/stdlib.h:62:
/Library/Developer/CommandLineTools/SDKs/MacOSX10.14.sdk/usr/include/sys/cdefs.h:784:2: error: Unsupported architecture
784 | <…

A look at Google's TPU evolution, its decision to sell chips to competitors after 12 years of internal use, the TPU vs. GPU architecture showdown, and more (The Chip Letter)
https://thechipletter.substack.com/p/tpu-mania

Has anyone made a good video on the obscure IBM System 9000?
It's a 68000-based architecture that can support "large amounts of RAM", the article brags about "as much 2 megabytes".
https://archive.org/details/byte-magazine-1984-09/page/n225/mode/2up

Designers are always inventing new titles. Web designer, UI designer, UX designer, UX/UI designer, Product designer, UX/AI designer (???)
These are all the same!
As long as your work is mediated entirely by Figma and Jira, you are just a production artist.
Fighting with PMs over who "gets" to do strategy (usually: neither of you) won't help you. Lean into the core #design

I blogged about my cute little open source computer, the #mntpocketreform, and how it is connected to the idea of #permacomputing.
https://

#Rotterdam centraal station

This is a truly fascinating use of LLMs. I hope we get to see this soon.
https://github.com/DGoettlich/history-llms
A family of 4 billion (B) parameter LLMs based on the Qwen3 architecture trained on 80B tokens of historical data up to knowledge-cutoffs : 1913, 1929, 1933, 1939…

Where am I?

Urban Spots ✴️
城市噪点 ✴️
📷 Nikon F4E
🎞️ Ilford HP5 Plus 400, expired 1993
#filmphotography #Photography #blackandwhite

Just kinda thinking out loud here, don't have a solution or architecture in mind: I really want a better solution for high speed FPGA / ASIC protocol design and debug.
In particular:
1) Using scopehal protocol captures as stimuli for a virtual DUT (i.e. decoded 8b10b of PCIe or something into RTL simulator, pretending to be a SERDES)

Revamped my domain name site #indieweb using

Series C, Episode 13 - Terminal
ZEN: The information you require has been stored via a coded route, and retrieval is not possible without the correct security command sequence.
CALLY: And that sequence is known only to Avon, I imagine.
ZEN: Confirmed.
https://blake.torpidity.net/m/313/100

In the last ~6 months someone linked to their blog about using JS-free custom elements for page structure and style in place of a methodology like BEM, e.g. using `<my-element>`/`my-element {}` instead of `<div class="my-element">`/`.my-element {}`. Might have referenced [HUG CSS](https://

For a few years I had an office in the Frank Gehry-designed Stata Center at MIT. Here are some photos I took on a sunny autumn day using a film camera in 2005. This building had a lot of problems but it definitely made a statement. #architecture #FrankGehry

@… thanks. I was happy to use poudriere in the past but now, I have a Kubuntu host.
FreeBSD-14.3-RELEASE-arm64-aarch64-disc1.iso
First attempt: I forgot the disclosure triangle when creating a new machine – allowed the wrong architecture. Oops.
A later attempt: the VM seemed to forget that the .iso had been specified.
Eve…

AheadComputing, which develops a breakthrough RISC-V CPU architecture for AI, raised a $30M Seed2 led by Eclipse, Toyota, and Cambium, for $53M in total funding (Kyt Dotson/SiliconANGLE)
https://siliconangle.com/2026/01/21/aheadcomputing-lands-3…

This white paper provides a comprehensive analysis of modern warfare through five interconnected characteristics that have been prominently displayed throughout the Ukraine conflict:
- The rise of autonomous systems and their impact on force architecture
- The information domain as a critical battleground
- Electronic warfare and spectrum superiority
- The challenges of sustaining logistics in contested environments
- The evolution of air defense strategy
…

Polyharmonic Cascade
Yuriy N. Bakhvalov
https://arxiv.org/abs/2512.17671 https://arxiv.org/pdf/2512.17671 https://arxiv.org/html/2512.17671
arXiv:2512.17671v1 Announce Type: new
Abstract: This paper presents a deep machine learning architecture, the "polyharmonic cascade" -- a sequence of packages of polyharmonic splines, where each layer is rigorously derived from the theory of random functions and the principles of indifference. This makes it possible to approximate nonlinear functions of arbitrary complexity while preserving global smoothness and a probabilistic interpretation. For the polyharmonic cascade, a training method alternative to gradient descent is proposed: instead of directly optimizing the coefficients, one solves a single global linear system on each batch with respect to the function values at fixed "constellations" of nodes. This yields synchronized updates of all layers, preserves the probabilistic interpretation of individual layers and theoretical consistency with the original model, and scales well: all computations reduce to 2D matrix operations efficiently executed on a GPU. Fast learning without overfitting on MNIST is demonstrated.
toXiv_bot_toot

The Architecture of the Venezuelan Crisis | by Ireneusz Tarnowski | Jan, 2026 | Medium
https://medium.com/@ireneusz.tarnowski/the-architecture-of-the-venezuelan-crisis-a9a1c3737bab

Nächster Talk beim @…: @… erzählt uns seine Ideen zu "Democratic Architecture"

As announced, I have tagged a prerelease yesterday that oofers the basic me_cleaner capabilities.
Today, I have documented some architecture and design considerations:
https://github.com/platform-system-interface/intel_fw/pull/77
Testers welcome. If you ru…

The coolest material thing I saw in 2025 (even if utterly impractical for now) was Iceland's entry in the Venice Architecture Biennale: the futuristic «lavaforming», using lava as a building material.
https://www.honnunarmidstod.is/en/feneyjatviaeringurinn/lavafor…

Medium-endian architecture

NeuroSketch: An Effective Framework for Neural Decoding via Systematic Architectural Optimization
Gaorui Zhang, Zhizhang Yuan, Jialan Yang, Junru Chen, Li Meng, Yang Yang
https://arxiv.org/abs/2512.09524 https://arxiv.org/pdf/2512.09524 https://arxiv.org/html/2512.09524
arXiv:2512.09524v1 Announce Type: new
Abstract: Neural decoding, a critical component of Brain-Computer Interface (BCI), has recently attracted increasing research interest. Previous research has focused on leveraging signal processing and deep learning methods to enhance neural decoding performance. However, the in-depth exploration of model architectures remains underexplored, despite its proven effectiveness in other tasks such as energy forecasting and image classification. In this study, we propose NeuroSketch, an effective framework for neural decoding via systematic architecture optimization. Starting with the basic architecture study, we find that CNN-2D outperforms other architectures in neural decoding tasks and explore its effectiveness from temporal and spatial perspectives. Building on this, we optimize the architecture from macro- to micro-level, achieving improvements in performance at each step. The exploration process and model validations take over 5,000 experiments spanning three distinct modalities (visual, auditory, and speech), three types of brain signals (EEG, SEEG, and ECoG), and eight diverse decoding tasks. Experimental results indicate that NeuroSketch achieves state-of-the-art (SOTA) performance across all evaluated datasets, positioning it as a powerful tool for neural decoding. Our code and scripts are available at https://github.com/Galaxy-Dawn/NeuroSketch.
toXiv_bot_toot

The National Archives and RIBA to partner https://www.nationalarchives.gov.uk/about/news/the-national-archives-and-riba-to-partner/

Moody Urbanity - ONE 1️⃣
情绪化城市 - 壹 1️⃣
📷 Nikon FE
🎞️ Fujifilm NEOPAN SS, expired 1993
#filmphotography #Photography #blackandwhite

Replaced article(s) found for cs.GR. https://arxiv.org/list/cs.GR/new
[1/1]:
- Controllable Video Generation: A Survey
Yue Ma, et al.
https://arxiv.org/abs/2507.16869 https://mastoxiv.page/@arXiv_csGR_bot/114907178598354130
- Lightning Fast Caching-based Parallel Denoising Prediction for Accelerating Talking Head Generation
Jianzhi Long, Wenhao Sun, Rongcheng Tu, Dacheng Tao
https://arxiv.org/abs/2509.00052 https://mastoxiv.page/@arXiv_csGR_bot/115139250819269869
- MimicKit: A Reinforcement Learning Framework for Motion Imitation and Control
Xue Bin Peng
https://arxiv.org/abs/2510.13794 https://mastoxiv.page/@arXiv_csGR_bot/115382726856686148
- TIDI-GS: Floater Suppression in 3D Gaussian Splatting for Enhanced Indoor Scene Fidelity
Sooyeun Yang, Cheyul Im, Jee Won Lee, Jongseong Brad Choi
https://arxiv.org/abs/2601.09291 https://mastoxiv.page/@arXiv_csGR_bot/115898204587831863
- Eye-tracked Virtual Reality: A Comprehensive Survey on Methods and Privacy Challenges
Bozkir, S\Ozdel, Wang, David-John, Gao, Butler, Jain, Kasneci
https://arxiv.org/abs/2305.14080
- Hi5: Synthetic Data for Inclusive, Robust, Hand Pose Estimation
Hasan, Ozel, Long, Martin, Potter, Adnan, Lee, Hoque
https://arxiv.org/abs/2406.03599 https://mastoxiv.page/@arXiv_csCV_bot/112573997027314918
- A Text-to-3D Framework for Joint Generation of CG-Ready Humans and Compatible Garments
Zhiyao Sun, Yu-Hui Wen, Ho-Jui Fang, Sheng Ye, Matthieu Lin, Tian Lv, Yong-Jin Liu
https://arxiv.org/abs/2503.12052 https://mastoxiv.page/@arXiv_csCV_bot/114182219370820263
- A Unified Architecture for N-Dimensional Visualization and Simulation: 4D Implementation and Eval...
Hirohito Arai
https://arxiv.org/abs/2512.01501 https://mastoxiv.page/@arXiv_csCG_bot/115648840470000746
toXiv_bot_toot

The Case Against #Microservices
https://sashafoundtherootcauseagain.substack.com/p/the-case-against-microservices

@… @… The whole south from Spain to Egypt have similar architecture - when we were just needed some place to spend a night and leave at the morning -cemetery's parkings are perfect spots.

When people say they need hope about climate change, what are they really asking for?
New piece breaks down why hope collapses so easily and what stronger feelings can replace it.
Not prediction. Participation.
Read it here: [link]
https://www.b…

Series C, Episode 05 - The Harvest of Kairos
AVON: Cally. [Grasps her by the arms]
TARRANT: [Enters] Is this the time or the place? That thing has warped your reason, Avon. It's even warped your notorious instinct for looking after number one. We are in danger, can't you understand that?
https://blake.torpidity.net/m/3…

from my link log —
Do you really need Redis? PostgreSQL does queuing, locking, and pub/sub.
https://spin.atomicobject.com/2021/02/04/redis-postgresql/
saved 2021-06-12

Looking at the ARMv8-A ARM and I see implementer ID 0x44 is DEC.
I'm... very confident DEC has never built an ARMv8 CPU core.
Is this just copy-pasted from earlier docs because DEC had an ARM architecture license in years past? Or is this license still valid and held by DEC's successors (I think DEC -> Compaq -> HP -> HPE, does HPE make custom ARM cores?)

Misplaced neurons reveal the brain's adaptability https://www.sflorg.com/2026/01/ns01162601.html Position-independent emergence of neocortical neuron molecular identity, connectivity and function

Moody Urbanity - Past II 📴
情绪化城市 - 过去 II 📴
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Ilford HP5, expired 1993
#filmphotography #Photography #blackandwhite

Vibe coding is the single biggest transformation since ChatGPT 3.5, it is one of the biggest transformations since the dawn of computing. It will thoroughly transform Open Source. We are not ready. #VibeCoding #OpenSource

Anthropic commits to buy $30B in Azure capacity in a new deal with Microsoft and Nvidia, which commit to invest up to $5B and $10B, respectively, in Anthropic (Microsoft)
https://blogs.microsoft.com/blog/2025/11/18/microsoft-nvidia-…

Kids on the bus, all playing video games walking to the back to find a seat... jeesh, teach them to read and write some Rust 🦀 or computer architecture and have them implement RISC-V cores or smth.

I Love You, #Redis, But I’m Leaving You for #SolidQueue
https://www.simplethread.com/redis-solidqueue/…

You Only Train Once: Differentiable Subset Selection for Omics Data
Daphn\'e Chopard, Jorge da Silva Gon\c{c}alves, Irene Cannistraci, Thomas M. Sutter, Julia E. Vogt
https://arxiv.org/abs/2512.17678 https://arxiv.org/pdf/2512.17678 https://arxiv.org/html/2512.17678
arXiv:2512.17678v1 Announce Type: new
Abstract: Selecting compact and informative gene subsets from single-cell transcriptomic data is essential for biomarker discovery, improving interpretability, and cost-effective profiling. However, most existing feature selection approaches either operate as multi-stage pipelines or rely on post hoc feature attribution, making selection and prediction weakly coupled. In this work, we present YOTO (you only train once), an end-to-end framework that jointly identifies discrete gene subsets and performs prediction within a single differentiable architecture. In our model, the prediction task directly guides which genes are selected, while the learned subsets, in turn, shape the predictive representation. This closed feedback loop enables the model to iteratively refine both what it selects and how it predicts during training. Unlike existing approaches, YOTO enforces sparsity so that only the selected genes contribute to inference, eliminating the need to train additional downstream classifiers. Through a multi-task learning design, the model learns shared representations across related objectives, allowing partially labeled datasets to inform one another, and discovering gene subsets that generalize across tasks without additional training steps. We evaluate YOTO on two representative single-cell RNA-seq datasets, showing that it consistently outperforms state-of-the-art baselines. These results demonstrate that sparse, end-to-end, multi-task gene subset selection improves predictive performance and yields compact and meaningful gene subsets, advancing biomarker discovery and single-cell analysis.
toXiv_bot_toot

City Spot ⭕️
城市点阵 ⭕️
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Lucky SHD 400
#filmphotography #Photography #blackandwhite

When you say you need hope about climate change, what do you really need?
Maybe it's proof your actions matter.
Maybe it's permission to want a better future.
Maybe it's feeling less alone.
Your brain treats hope like a prediction. When reality doesn't match, hope breaks.
But agency doesn't break.
Community doesn't break.
Imagination doesn't break.
What if we stopped chasing hope and built something stronger?
Read …

Trump was sued on Friday by preservationists
asking a federal court to halt his White House ballroom project until it goes through multiple independent reviews and wins approval from Congress.
The National Trust for Historic Preservation, a privately funded group,
is asking the U.S. District Court to block Trump’s White House ballroom project,
which already has involved razing the East Wing,
until it goes through comprehensive design reviews, environmental asses…

If you are up this late, maybe you want to stay up later. Go ahead, read my latest essay, What Did Vibe Coding just do to the Commons?
https://varnelis.net/works_and_projects/what-did-vibe-coding-just-do-to-the-commons/

The Allen Institute of AI launches Bolmo 7B and Bolmo 1B, claiming they are "the first fully open byte-level language models", built on its Olmo 3 models (Emilia David/VentureBeat)
https://venturebeat.com/ai/bolmos-architecture-unlock…

Spatially-informed transformers: Injecting geostatistical covariance biases into self-attention for spatio-temporal forecasting
Yuri Calleo
https://arxiv.org/abs/2512.17696 https://arxiv.org/pdf/2512.17696 https://arxiv.org/html/2512.17696
arXiv:2512.17696v1 Announce Type: new
Abstract: The modeling of high-dimensional spatio-temporal processes presents a fundamental dichotomy between the probabilistic rigor of classical geostatistics and the flexible, high-capacity representations of deep learning. While Gaussian processes offer theoretical consistency and exact uncertainty quantification, their prohibitive computational scaling renders them impractical for massive sensor networks. Conversely, modern transformer architectures excel at sequence modeling but inherently lack a geometric inductive bias, treating spatial sensors as permutation-invariant tokens without a native understanding of distance. In this work, we propose a spatially-informed transformer, a hybrid architecture that injects a geostatistical inductive bias directly into the self-attention mechanism via a learnable covariance kernel. By formally decomposing the attention structure into a stationary physical prior and a non-stationary data-driven residual, we impose a soft topological constraint that favors spatially proximal interactions while retaining the capacity to model complex dynamics. We demonstrate the phenomenon of ``Deep Variography'', where the network successfully recovers the true spatial decay parameters of the underlying process end-to-end via backpropagation. Extensive experiments on synthetic Gaussian random fields and real-world traffic benchmarks confirm that our method outperforms state-of-the-art graph neural networks. Furthermore, rigorous statistical validation confirms that the proposed method delivers not only superior predictive accuracy but also well-calibrated probabilistic forecasts, effectively bridging the gap between physics-aware modeling and data-driven learning.
toXiv_bot_toot

Nvidia launches Nemotron 3, a family of AI models using a hybrid mixture-of-experts architecture and the Mamba-Transformer design, in 30B, 100B, and ~500B sizes (Emilia David/VentureBeat)
https://venturebeat.com/ai/nvidia-debuts-nemotron-3-w…

How Qwen 3 outcompetes OpenAI and DeepSeek https://apidog.com/blog/qwen-3-outcompetes-openai-and-deepseek/ "Qwen 3 shines with its variety of model sizes", "small models for edge devices or massive ones for heavy lifting";

As I'm grading #JCON's speaker proposals, I noticed an interesting behavior of mine. For Java proposals, I don't take much time in evaluating. The title might even be enough; architecture proposals take much more time as I go back and forth. I wonder why.

DeepSeek researchers detail a new mHC architecture they used to train 3B, 9B, and 27B models, finding it scaled without adding significant computational burden (Vincent Chow/South China Morning Post)
https://www.scmp.com/tech/big-tech/article

Crosslisted article(s) found for cs.LG. https://arxiv.org/list/cs.LG/new
[3/3]:
- Fraud detection in credit card transactions using Quantum-Assisted Restricted Boltzmann Machines
Jo\~ao Marcos Cavalcanti de Albuquerque Neto, Gustavo Castro do Amaral, Guilherme Penello Tempor\~ao
https://arxiv.org/abs/2512.17660 https://mastoxiv.page/@arXiv_quantph_bot/115762703945731580
- Vidarc: Embodied Video Diffusion Model for Closed-loop Control
Feng, Xiang, Mao, Tan, Zhang, Huang, Zheng, Liu, Su, Zhu
https://arxiv.org/abs/2512.17661 https://mastoxiv.page/@arXiv_csRO_bot/115762650859932523
- Imputation Uncertainty in Interpretable Machine Learning Methods
Pegah Golchian, Marvin N. Wright
https://arxiv.org/abs/2512.17689 https://mastoxiv.page/@arXiv_statML_bot/115762577479255577
- Revisiting the Broken Symmetry Phase of Solid Hydrogen: A Neural Network Variational Monte Carlo ...
Shengdu Chai, Chen Lin, Xinyang Dong, Yuqiang Li, Wanli Ouyang, Lei Wang, X. C. Xie
https://arxiv.org/abs/2512.17703 https://mastoxiv.page/@arXiv_condmatstrel_bot/115762481116668454
- Breast Cancer Neoadjuvant Chemotherapy Treatment Response Prediction Using Aligned Longitudinal M...
Rahul Ravi, Ruizhe Li, Tarek Abdelfatah, Stephen Chan, Xin Chen
https://arxiv.org/abs/2512.17759 https://mastoxiv.page/@arXiv_eessIV_bot/115762481771898369
- MedNeXt-v2: Scaling 3D ConvNeXts for Large-Scale Supervised Representation Learning in Medical Im...
Roy, Kirchhoff, Ulrich, Rokuss, Wald, Isensee, Maier-Hein
https://arxiv.org/abs/2512.17774 https://mastoxiv.page/@arXiv_eessIV_bot/115762492258209812
- Domain-Aware Quantum Circuit for QML
Gurinder Singh, Thaddeus Pellegrini, Kenneth M. Merz, Jr
https://arxiv.org/abs/2512.17800 https://mastoxiv.page/@arXiv_quantph_bot/115762723607200478
- Visually Prompted Benchmarks Are Surprisingly Fragile
Feng, Lian, Dunlap, Shu, Wang, Wang, Darrell, Suhr, Kanazawa
https://arxiv.org/abs/2512.17875 https://mastoxiv.page/@arXiv_csCV_bot/115762781936221554
- Learning vertical coordinates via automatic differentiation of a dynamical core
Tim Whittaker, Seth Taylor, Elsa Cardoso-Bihlo, Alejandro Di Luca, Alex Bihlo
https://arxiv.org/abs/2512.17877 https://mastoxiv.page/@arXiv_physicsaoph_bot/115762405092703069
- RadarGen: Automotive Radar Point Cloud Generation from Cameras
Tomer Borreda, Fangqiang Ding, Sanja Fidler, Shengyu Huang, Or Litany
https://arxiv.org/abs/2512.17897 https://mastoxiv.page/@arXiv_csCV_bot/115762783246540528
- Distributionally Robust Imitation Learning: Layered Control Architecture for Certifiable Autonomy
Gahlawat, Aboudonia, Banik, Hovakimyan, Matni, Ames, Zardini, Speranzon
https://arxiv.org/abs/2512.17899 https://mastoxiv.page/@arXiv_eessSY_bot/115762532257741954
- Re-Depth Anything: Test-Time Depth Refinement via Self-Supervised Re-lighting
Ananta R. Bhattarai, Helge Rhodin
https://arxiv.org/abs/2512.17908 https://mastoxiv.page/@arXiv_csCV_bot/115762785868778349
toXiv_bot_toot

Moody Urbanity - TWO II 2️⃣
情绪化城市 - 贰 II 2️⃣
📷 Nikon FE
🎞️ Fujifilm NEOPAN SS, expired 1993
#filmphotography #Photography #blackandwhite

The most celebrated architect of his age, a wry, pugnacious, singular genius,
Frank Gehry, died Friday at 96
In Bilbao and Berlin, New York and, especially, Los Angeles
— where the Canadian-born Gehry settled and became inseparable from the city
— he left behind playful, polarizing, materially and technologically stunning buildings

Series D, Episode 11 - Orbit
SERVALAN: Don't concern yourself, Egrorian. I encourage ambition.
EGRORIAN: I never conspired against you, I swear it! Orac, acknowledge my instructions. [Orac buzzes, but says nothing.] I don't understand.
https://blake.torpidity.net/m/411/363 B7B3…

US CHILDREN II 🧒
我們這些孩童 II 🧒
📷 Zeiss IKON Super Ikonta 533/16
🎞️ Lucky SHD 400
#filmphotography #Photography #blackandwhite

Majestic Labs, which makes patent-pending server architecture that promises more memory capacity, raised $100M, including a $71M Series A led by Bow Wave (Jennifer Elias/CNBC)
https://www.cnbc.com/2025/11/10/majestic-labs-meta-google-ai-funding.html
…

Moody Urbanity - Old & New V 🔆
情绪化城市 - 新与旧 V 🔆
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Ilford HP5 Plus 400, expired 1993
#filmphotography #Photography #blackandwhite

Metropolitana III🔶
城 III🔶
📷 Pentax MX
🎞️ Ilford Pan 100
#filmphotography #Photography #blackandwhite

How Pathway, a startup developing an alternative to the transformer, aims to use its Dragon Hatchling architecture to create a new class of adaptive AI systems (Steven Rosenbush/Wall Street Journal)
https://www.wsj.com/articles/an-ai…

Un Momento 🕰️
一瞬 🕰️
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Lucky SHD 400
#filmphotography #Photography #blackandwhite

A technical deep dive into Amazon's Trainium3 accelerator, including its server SKUs' specifications, silicon design, power budget, and bill of materials (SemiAnalysis)
https://newsletter.semianalysis.com/p/aws-trainium3-deep-dive-a-potential

Google debuts Titans, an architecture combining RNN speed with transformer performance for real-time learning, able to scale effectively to a 2M context window (Google Research)
https://research.google/blog/titans-miras-helping-ai-have-long-term-memory/

Urban Adventure 🏞️
城市探险 🏞️
📷Nikon FE
🎞️Fuijfilm NEOPAN SS, expired 1993
#filmphotography #Photography #blackandwhite

Qualcomm unveils the Dragonwing IQ10 Series, a full-stack robotics architecture integrating hardware, software, and AI for industrial and consumer humanoids (Katie Collins/CNET)
https://www.cnet.com/tech/qualcomm-announces-dragonwing-iq10-…

Moody Urbanity - Old & New VI 🔆
情绪化城市 - 新与旧 VI 🔆
📷 Zeiss Ikon Super Ikonta 533/16
🎞️ Ilford HP5 Plus 400, expired 1993
#filmphotography #Photography #blackandwhite

Nvidia launches the Vera Rubin platform, saying it will offer dramatic reductions in inference and training costs compared to Blackwell, across six new chips (Stevie Bonifield/The Verge)
https://www.theverge.com/tech/855412/nvidia-launches-ve…

Qualcomm unveils Snapdragon X2 Plus chips for PCs, available in 10-core and 6-core variants, manufactured on TSMC's N3P node using Qualcomm's Oryon architecture (Jake Roach/Tom's Hardware)
https://www.tomshar…

Metropolitana VI - Asymmetry ✅
城 VI - 非对称 ✅
📷 Pentax MX
🎞️ Ilford Pan 100
#filmphotography #Photography #blackandwhite

A deep dive into co-packaged optics, long promised to transform data center connectivity, covering benefits, challenges, architecture, key companies, and more (Dylan Patel/SemiAnalysis)
https://newsletter.semianalysis.com/p/co-packaged-optics-cpo-book-scaling…

Moody Urbanity - Relations VI 🧬
情绪化城市 - 关系 VI 🧬
📷 Zeiss IKON Super Ikonta 533/16
🎞️ Ilford HP5 400, expired 1993
#filmphotography #Photography #blackandwhite

Gloom & Bloom ☁️🌺
黑暗绽放 ☁️ 🌺
📷 Zeiss IKON Super Ikonta 533/16
🎞️ Lucky SHD 400
#filmphotography #Photography #blackandwhite