Tootfinder

#Tesla-Fans scheitern bei #Autopilot-Testfahrt nach 96 Kilometern
https://www.

🇺🇦 #NowPlaying on KEXP's #DriveTime
The Staple Singers:
🎵 The Weight
#TheStapleSingers
https://thestaplesingers.bandcamp.com/track/the-weight-6
https://open.spotify.com/track/00bOhb4584JjyfTiXX81mO

GIFT: Gradient-aware Immunization of diffusion models against malicious Fine-Tuning with safe concepts retention
Amro Abdalla, Ismail Shaheen, Dan DeGenaro, Rupayan Mallick, Bogdan Raita, Sarah Adel Bargal
https://arxiv.org/abs/2507.13598

GIFT: Gradient-aware Immunization of diffusion models against malicious Fine-Tuning with safe concepts retention
We present \textbf{GIFT}: a \textbf{G}radient-aware \textbf{I}mmunization technique to defend diffusion models against malicious \textbf{F}ine-\textbf{T}uning while preserving their ability to generate safe content. Existing safety mechanisms like safety checkers are easily bypassed, and concept erasure methods fail under adversarial fine-tuning. GIFT addresses this by framing immunization as a bi-level optimization problem: the upper-level objective degrades the model's ability to represent ha…

from my link log —
Why some of us like "interdiff" code review.
https://gist.github.com/thoughtpolice/9c45287550a56b2047c6311fbadebed2
saved 2025-08-20

DiEP: Adaptive Mixture-of-Experts Compression through Differentiable Expert Pruning
Sikai Bai, Haoxi Li, Jie Zhang, Zicong Hong, Song Guo
https://arxiv.org/abs/2509.16105 https:…

DiEP: Adaptive Mixture-of-Experts Compression through Differentiable Expert Pruning
Despite the significant breakthrough of Mixture-of-Experts (MoE), the increasing scale of these MoE models presents huge memory and storage challenges. Existing MoE pruning methods, which involve reducing parameter size with a uniform sparsity across all layers, often lead to suboptimal outcomes and performance degradation due to varying expert redundancy in different MoE layers. To address this, we propose a non-uniform pruning strategy, dubbed \textbf{Di}fferentiable \textbf{E}xpert \textbf{P…

« Adult human cortex does not reorganize after amputation » (and certainly also for other species). So, do not trust textbooks (or LLMs that statistically parameterize those). #brain #plasticity #neuroscience

Adult human cortex does not reorganize after amputation
The results from a new longitudinal study contradict classic findings in monkeys but may not warrant a rewriting of the textbooks just yet.

"Akademische Textverarbeitung: Digitale Literalität und offene Formate in Bibliotheken" https://doi.org/10.1515/abitech-2025-0043

Montag: Erlaubnis für nächsten Starship-Testflug, Zugkupplung mit Ultrabreitband
Riesenrakete vor nächstem Start UWB für mehr Züge im Netz ÖPNV im weltweiten Vergleich Schweizer Überwachung vertreibt Firmen Microsofts Preiserhöhung

We are taking Xogot for a spin on M-based Macs using Mac Catalyst.
You can download now from TestFlight, or you can sign up here: https://beta.xogot.com

On weighted multilinear polynomial averages in finite fields
Guo-Dong Hong
https://arxiv.org/abs/2507.14414 https://arxiv.org/pdf/250…

On weighted multilinear polynomial averages in finite fields
We study the weighted multilinear polynomial averages in finite fields. The essential ingredient is the $u^s$-norm control of the corresponding weighted multilinear polynomial averages in finite fields, which is motivated by Teräväinen \cite{T24}. As an application, we prove an asymptotic formula for the number of the following multidimensional rational function progressions in the subsets of $\mathbb{F}_p^D$: \[ \textbf{x}, \textbf{x}+ P_1(φ(y))v_1,\cdots, \textbf{x}+ P_k(φ(y))v_k, \] wher…