Tootfinder

I have taken to looking for Copilot AI in GitHub responses before filing issues. If it’s there, I don’t file. Feeling a bit validated here:
https://hails.org/@hailey/114752144098708214
(limited alt to come, but if you want more ask OP…)

Hailey (@hailey@hails.org)
Attached: 2 images Sent a pull request to Audacity fixing a crash bug I'd been running into frequently. The cause was an out-of-bounds memmove. Classic C++ areas. Anyway I got a fucking copilot review on my PR which left two comments, both completely wrong, one of which suggesting I reintroduce the out of bounds memory access. I'm furious!

UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
Zihao Huang, Yu Bao, Qiyang Min, Siyan Chen, Ran Guo, Hongzhi Huang, Defa Zhu, Yutao Zeng, Banggu Wu, Xun Zhou, Siyuan Qiao
https://arxiv.org/abs/2508.18756

UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning
While Mixture of Experts (MoE) models achieve remarkable efficiency by activating only subsets of parameters, they suffer from high memory access costs during inference. Memory-layer architectures offer an appealing alternative with very few memory access, but previous attempts like UltraMem have only matched the performance of 2-expert MoE models, falling significantly short of state-of-the-art 8-expert configurations. We present UltraMemV2, a redesigned memory-layer architecture that closes t…

Magnetization switching by current in an elemental ferromagnetic single layer
Yanyan Yang, Weiwei Lin
https://arxiv.org/abs/2507.22016 https://arxiv.org/pd…

Magnetization switching by current in an elemental ferromagnetic single layer
Current-induced magnetization switching, a fundamental phenomenon related to spin-transport of electrons, enables non-voltaic and fast information write, facilitating applications in low-power memory and logic devices. However, magnetization switching by spin-orbit torques is usually attributed to current flowing in the nonmagnetic metal layer of multilayers or in magnetic alloys with heavy elements. Here, we report perpendicular magnetization switching induced by current flowing in an elementa…

The Persistence of Nonlinear Gravitational Wave Memory
Robert R. Caldwell
https://arxiv.org/abs/2506.20751 https://arxiv.org/pdf/2506…

The Persistence of Nonlinear Gravitational Wave Memory
Nonlinear gravitational wave memory is a surprise of theoretical physics. Whereas it is understood that a gravitational wave induces oscillatory squeezing and stretching motion in a collection of freely-falling test masses, it is unexpected that the wave leaves a residual displacement of the test masses. This displacement is the tribute in memoriam to the passing wave. The memory originates in a nonlinear feature of gravitation. Whilst merging black holes are a significant source of gravitation…

StruMamba3D: Exploring Structural Mamba for Self-supervised Point Cloud Representation Learning
Chuxin Wang, Yixin Zha, Wenfei Yang, Tianzhu Zhang
https://arxiv.org/abs/2506.21541

StruMamba3D: Exploring Structural Mamba for Self-supervised Point Cloud Representation Learning
Recently, Mamba-based methods have demonstrated impressive performance in point cloud representation learning by leveraging State Space Model (SSM) with the efficient context modeling ability and linear complexity. However, these methods still face two key issues that limit the potential of SSM: Destroying the adjacency of 3D points during SSM processing and failing to retain long-sequence memory as the input length increases in downstream tasks. To address these issues, we propose StruMamba3D,…

Thermal nature of the causal diamond horizon: A hidden property of the inertial propagator
Nada Eissa, Carlos R. Ord\'o\~nez, Gustavo Valdivia-Mera
https://arxiv.org/abs/2508.16880

Thermal nature of the causal diamond horizon: A hidden property of the inertial propagator
Inspired by the novel idea proposed by T.~Padmanabhan in Phys. Rev. D 100, 045024 (2019), we develop a method to uncover the hidden thermal properties of the inertial Feynman propagator in Minkowski spacetime in a causally consistent manner. This, in turn, enables a coherent interpretation based on future-directed propagation. In this method, the Fourier transform is implemented following the convention used in the analysis of vacuum fluctuations. As a result, future-directed propagation across…

A Programming Language for Feasible Solutions
Weijun Chen, Yuxi Fu, Huan Long
https://arxiv.org/abs/2507.19176 https://arxiv.org/pdf/2507.19176

A Programming Language for Feasible Solutions
Runtime efficiency and termination are crucial properties in the studies of program verification. Instead of dealing with these issues in an ad hoc manner, it would be useful to develop a robust framework in which such properties are guaranteed by design. This paper introduces a new imperative programming language whose design is grounded in a static type system that ensures the following equivalence property: All definable programs are guaranteed to run in polynomial time; Conversely, all prob…

Stretched Brownian Motion: convergence of dual optimising sequences
Walter Schachermayer, Pietro Siorpaes
https://arxiv.org/abs/2508.20017 https://arxiv.or…

Stretched Brownian Motion: convergence of dual optimising sequences
We consider an irreducible pair $μ\leq_c ν$ of probability measures on $\mathbb{R}^d$ in convex order. In arXiv:2306.11019, Backhoff, Beiglböck, Schachermayer and Tschiderer have shown that the Stretched Brownian Motion from $μ$ to $ν$ is a Bass martingale, that there exists a dual optimiser $ψ_{lim}$, and the following somewhat surprising convergence result: by adding affine functions, one can make any dual optimising sequence $(ψ_n)_n$ (satisfying some minor technical conditions) conve…

I love the fact that switching out a failing SSD just involved formatting the new SSD, copying the files over via `rsync`, powering off the computer, and swapping the new SSD in for the old one. I did have to make a minor tweak to the `fstab` because I decided to go with XFS, but the down time for the computer was only for as long as it took me to get the SSD swapped in.
#linux

Kinetic pathways of coesite densification from metadynamics
David Vrba, Roman Marto\v{n}\'ak
https://arxiv.org/abs/2508.19716 https://arxiv.org/pdf/250…

Kinetic pathways of coesite densification from metadynamics
We study compression of coesite to pressures above 35 GPa, substantially beyond the equilibrium transition pressure to octahedral phases (8 GPa to stishovite). Experiments at room temperature showed that up to 30 GPa the metastable coesite structure develops only minor displacive changes (coesite-II and coesite-III) while the Si atoms remain 4-coordinated. Beyond 30 GPa, reconstructive transformations start, following different pathways from the complex structure of coesite. Besides amorphizati…