Tootfinder

"As a result, Europe is now 2.5 °C hotter than before the mass burning of fossil fuels, almost double the global average 1.4 °C rise. Temperatures in Europe have risen by 0.56 °C a decade since the mid-1990s, compared with 0.42 °C in North America, 0.36 °C in Africa and 0.27 °C in South America."
(Edit: Degree symbols added for readability.)

Growing geopolitical competition is accompanying environmental disruption.
Countries including Russia, China, and the United States are investing in the Arctic
and consolidating their interests in strategic access, resource exploitation, and alliance-building.
 
The consequences of this emerging Arctic Anthropocene are truly global
—from rising sea levels due to melting glaciers
to tensions between great powers determined to protect their territory and resou…

Unfrozen
A vital account of the state of the Arctic today—emphasising the twin dangers of climate change and geopolitical competition   Nowhere is the dual threat ...

#WritersCoffeeClub 10. Do you have a guiding principle as a writer? What is it? (Or why not?)
I don't think I've consciously formulated a guiding principle, but looking at how I write, perhaps it's this: trust the reader.
I don't want to explain everything, tell people what to feel, or insist they interpret a character the way I do. And if I can make them laugh …

ProtoGIB-Workload: Learning Workload-Specific Neural Topology Prototypes across Subjects
Yuzhe Zhang, Yixi Zhang, Shengdian Jiang, Chengxi Xie, Jihong Wang, Huan Liu, Man Yao, Minnan Luo, Chao Shen
https://arxiv.org/abs/2608.10647 https://arxiv.org/pdf/2608.10647 https://arxiv.org/html/2608.10647
arXiv:2608.10647v1 Announce Type: new
Abstract: Reliable electroencephalography (EEG)-based mental workload recognition is crucial for adaptive human-centered systems, yet practical deployment requires models to generalize to users unseen during training. Although functional connectivity graphs are widely adopted to capture workload-related neural interactions, they inherently entangle task-relevant structures with subject-specific physiological traits and sample-level noise. This entanglement often leads models to learn structural shortcuts, severely degrading cross-subject generalization. To address this, we propose ProtoGIB-Workload, a novel framework that explicitly regularizes and aligns graph structures for subject-independent workload recognition. Our approach introduces a Stochastic Graph Information Bottleneck (SGIB) to compress dense correlation priors into compact, task-relevant subgraphs, filtering out input-related redundancy. Crucially, to prevent the retention of subject-specific spurious edges, we propose a Class-Conditional Topology Stabilizer (CTS). Leveraging the fixed electrode coordinates of EEG data, CTS operates directly on graph-generation probabilities to encourage consistent edge-generation statistics across different subjects sharing the same workload class. Extensive experiments on two public EEG workload datasets and one in-house EEG cognitive load dataset of air traffic controllers under strict leave-one-subject-out (LOSO) protocols demonstrate that ProtoGIB-Workload significantly outperforms state-of-the-art temporal and graph-based baselines, improving the cross-subject Macro-F1 score by an average of 5.15% (up to 6.34%). Further analyses confirm that our method successfully extracts stable, cross-subject consistent neural connectivity patterns.
toXiv_bot_toot

i live in a small apartment without a lot of storage space. for LPs, i have two three-foot shelves, one for jazz, the other for everything else. last week, i made a new resolution to play at least one non-jazz record every day. here's how it's going.
#vinyl #OldVinyl

Let's actually talk about the "AI destroying the world" thing. One specific apocalypse scenario is the "paperclip maximizer" problem. The scenario goes that a super intelligence is tasked with maximizing the number of paperclips produced by a company. Because of a thing called "instrumental convergence," the AI kills all humans to prevent them from shutting it down, then it destroys everything else strip mining the Earth, etc.
Basically, the intelligence doesn't understand the actual goal of helping people hold paper together and therefore calculates intermediate goals that violate the *implicit* goal in order to fulfill the *explicit* goal.
The thing is, that's just capitalism. The whole AI scam is an example of this, where maximization of quarterly profit is an explicit goal and "destroying human civilization and most life on Earth through climate change" is a side effect of the "build as many datacenters as possible" subgoal.
Capitalism has been an apocalypse manifesting malicious intelligence since we figured out that burning fossil fuels changes the climate, back in 1824.

from my link log —
Error sets with eros: the missing piece in Rust error handling.
https://mcmah309.github.io/posts/the-missing-piece-in-rust-error-handling/
saved 2026-10-08

Countable Functionally Balanced Groups
Li-Hong Xie, Jiang Yang
https://arxiv.org/abs/2608.12494 https://arxiv.org/pdf/2608.12494 https://arxiv.org/html/2608.12494
arXiv:2608.12494v1 Announce Type: new
Abstract: We prove that every countable Hausdorff functionally balanced topological group is balanced, answering Question~13 in the problem list of Bouziad and Troallic. The proof does not assume metrizability, first countability, sequentiality, or a non-Archimedean local base. Its non-locally-precompact part combines an infinite-fibre amplification, a simultaneous Bernoulli colouring lemma for countably many arbitrary self-maps, a McShane extension with respect to a continuous left-invariant pseudometric, and the Protasov--Saryev criterion. We also prove a countable-union form of Question~4: in an \(\FSIN\) group, every left uniformly discrete family whose union is countable is right uniformly discrete. Consequently, Questions~7 and~8 have positive answers for \(\aleph_0\)-bounded \(\FSIN\) groups. We reconstruct all thirteen questions and their logical relations. The original problem chapter states, without proof, that a positive answer to Question~13 would imply Question~4 for all \(\aleph_0\)-bounded groups. We isolate the ambient reflection principle needed for that transfer and explain why neither subgroup reduction nor the present colouring proof establishes it. Therefore Question~12 is \emph{not} claimed as solved here: the precise boundary between the proved countable theorem and the asserted \(\aleph_0\)-bounded transfer is recorded explicitly.
toXiv_bot_toot

Spec Sheets Are Not Kernels: An ISA- and Source-Level Audit of INT8 Availability on NVIDIA Blackwell Ultra
Teng-Ruei Chen
https://arxiv.org/abs/2608.11693 https://arxiv.org/pdf/2608.11693 https://arxiv.org/html/2608.11693
arXiv:2608.11693v1 Announce Type: new
Abstract: NVIDIA's published specifications give the Blackwell Ultra GPU (B300) a dense-compute ratio of roughly 30:1 between FP8 and INT8 tensor-core throughput; its predecessors, H200 and B200, both provide 1:1. We audit what this deprioritization means in practice by tracing INT8 W8A8 support through four layers of the stack: the published specifications, the PTX ISA, NVIDIA's CUTLASS kernel library, and the two major open-source LLM serving engines (vLLM and SGLang). We find a consistent, layered withdrawal: (i) the PTX ISA never exposes the fifth-generation tensor-core integer path (tcgen05.mma with .kind::i8) on sm_103a, even though the same PTX revision extends the FP4 kinds to that target, leaving legacy warp-level IMMA as the only architecturally legal integer tensor-core path on B300; (ii) CUTLASS's kernel generator explicitly skips INT8 UMMA generation for any build targeting 103a, while generating FP8 unconditionally; (iii) vLLM ships no INT8 GEMM for Blackwell and fails with a hard runtime error at the first forward pass, after the model has loaded; and (iv) SGLang's ahead-of-time INT8 GEMM stops at Sm90, while its FP8 tuning configurations already cover B200. We document an escape hatch (rerouting vLLM's INT8 path to a JIT-compiled Triton backend via an environment variable), a false-negative trap in the obvious profiler methodology for detecting "native INT8" on sm_103, and the practical failure semantics that make naive testing expensive. Together, these findings show that a quantization format's availability is a property of the whole stack rather than of the model or the spec sheet. Four distinct layers, three of them NVIDIA's own, withdrew INT8 support in mutually consistent ways, and a format that is nominally present on the datasheet is, by default, undeployable on this hardware.
toXiv_bot_toot

Domestic politics severely constrain Germany’s options following the formal attribution of an explosive-laden Russian drone discovered near a Ukrainian cargo plane at Leipzig/Halle Airport.
Despite Chancellor Friedrich Merz and senior officials like Interior Minister Alexander Dobrindt stating that Berlin faces a daily campaign of Russian hybrid warfare,
Germany's retaliatory measures have been notably modest
—primarily consisting of diplomatic maneuvers such as closing…