Tootfinder

On my way to The Hague, to draw the red line for Dutch government with respect to Israel's continuing genocide in Gaza.
After last time's 100,000 participants, NS put in a lot of extra trains. My very early one already has a lot of passengers in red.
#rodelijn

GenBreak: Red Teaming Text-to-Image Generators Using Large Language Models
Zilong Wang, Xiang Zheng, Xiaosen Wang, Bo Wang, Xingjun Ma, Yu-Gang Jiang
https://arxiv.org/abs/2506.10047

GenBreak: Red Teaming Text-to-Image Generators Using Large Language Models
Text-to-image (T2I) models such as Stable Diffusion have advanced rapidly and are now widely used in content creation. However, these models can be misused to generate harmful content, including nudity or violence, posing significant safety risks. While most platforms employ content moderation systems, underlying vulnerabilities can still be exploited by determined adversaries. Recent research on red-teaming and adversarial attacks against T2I models has notable limitations: some studies succes…

Migrated two utility VMs to LXC containers. It's cool and all, and pretty neat that I can literally dump the filesystem from the VM and start it almost unmodified as a container, but udev rules don't work inside LXC so I'm kinda shafted with one of them. 50% is still a passing grade these days, right?

Dear old Violet.
She'll be turning 13 in October, so she's kind of on borrowed time for a greyhound. She seems to have lost most of her vision now and is underweight, though her appetite is still huge and she still loves her walks.
She was once a terror for the local cats and squirrels but now if they remain still she doesn't spot them. She still sniffs every place that she remembers them to have been.

Intrinsic Fingerprint of LLMs: Continue Training is NOT All You Need to Steal A Model!
Do-hyeon Yoon, Minsoo Chun, Thomas Allen, Hans M\"uller, Min Wang, Rajesh Sharma
https://arxiv.org/abs/2507.03014

Intrinsic Fingerprint of LLMs: Continue Training is NOT All You Need to Steal A Model!
Large language models (LLMs) face significant copyright and intellectual property challenges as the cost of training increases and model reuse becomes prevalent. While watermarking techniques have been proposed to protect model ownership, they may not be robust to continue training and development, posing serious threats to model attribution and copyright protection. This work introduces a simple yet effective approach for robust LLM fingerprinting based on intrinsic model characteristics. We d…

This https://arxiv.org/abs/2406.14200 has been replaced.
initial toot: https://mastoxiv.page/@arX…

Gaussian approximation of dynamic cavity equations for linearly-coupled stochastic dynamics
Stochastic dynamics on sparse graphs and disordered systems often lead to complex behaviors characterized by heterogeneity in time and spatial scales, slow relaxation, localization, and aging phenomena. The mathematical tools and approximation techniques required to analyze these complex systems are still under development, posing significant technical challenges and resulting in a reliance on numerical simulations. We introduce a novel computational framework for investigating the dynamics of …

Join Radu Pop and Pietro Mele at Berlin Buzzwords as they discuss building an extensible hybrid search solution with Elasticsearch. They will cover functional modeling, cluster architecture, and practical insights on managing billions of vectors in real-world scenarios. Radu and Pietro will also address hybrid reranking challenges and the limitations of standard fusion techniques, explaining their innovative approach.
Learn more: