If you have browsed my blog you may have noticed that each article has a different square image. A pattern of cubes that resembles a QR code but is not one. Colors that change from one post to another. And in the center, always the same kanji: 忍. Those images were not manually designed. They were generated by the content itself of each article. How it works When I publish an article the system takes two pieces of […]
arXiv:2603.12433v2 Announce Type: new Abstract: Model stitching, connecting early layers of one model (source) to later layers of another (target) via a light stitch layer, has served as a probe of representational compatibility. Prior work finds that models trained on the same dataset remain stitchable (negligible accuracy drop) despite different initializations or objectives. We revisit stitching for Vision Foundation Models (VFMs) that vary in objectives, data, and modality mix (e.g., CLIP, DINOv2, SigLIP 2) and ask: Are heterogeneous […]
arXiv:2602.16747v1 Announce Type: new Abstract: The reliability of medical LLM evaluation is critically undermined by data contamination and knowledge obsolescence, leading to inflated scores on static benchmarks. To address these challenges, we introduce LiveClin, a live benchmark designed for approximating real-world clinical practice. Built from contemporary, peer-reviewed case reports and updated biannually, LiveClin ensures clinical currency and resists data contamination. Using a verified AI-human workflow involving 239 physicians, we transform authentic patient cases into complex, multimodal evaluation scenarios […]
arXiv:2601.02531v1 Announce Type: new Abstract: Cooking recipes are complex procedures that require not only a fluent and factual text, but also accurate timing, temperature, and procedural coherence, as well as the correct composition of ingredients. Standard training procedures are primarily based on cross-entropy and focus solely on fluency. Building on RECIPE-NLG, we investigate the use of several composite objectives and present a new topological loss that represents ingredient lists as point clouds in embedding space, minimizing the divergence […]
arXiv:2604.06274v1 Announce Type: new Abstract: In recent years, the pace of development of information technology in various areas has increased drastically, forcing cybersecurity specialists to constantly review existing processes in order to prevent unauthorized access to confidential information. Using Ukraine as a primary case study, this paper explores the integration of international best practices, specifically ISO/IEC 27001 and the NIST Cybersecurity Framework, into national regulatory systems. A focus is placed on the transition from traditional compliance models to […]
Chevrolet provided flights from Albany, New York, to Las Vegas, Nevada, and accommodation so Ars could check out the new Grand Sport. Ars does not accept paid editorial content. Chevrolet has developed something of a modern tradition with recent generations of the Corvette: As a new generation approaches, the company rolls out the Grand Sport. It’s intended to be a sort of “sweet spot” version of the ’Vette, pairing the go-fast bits of the higher-spec machines with the […]
arXiv:2602.09235v1 Announce Type: new Abstract: Statistical data anonymization increasingly relies on fully synthetic microdata, for which classical identity disclosure measures are less informative than an adversary’s ability to infer sensitive attributes from released data. We introduce RAPID (Risk of Attribute Prediction–Induced Disclosure), a disclosure risk measure that directly quantifies inferential vulnerability under a realistic attack model. An adversary trains a predictive model solely on the released synthetic data and applies it to real individuals’ quasi-identifiers. For continuous sensitive […]
TL;DR Dead neurons silently waste compute and reduce effective model capacity in foundation models. Simple visualizations of the activation frequency make neuron health measurable. Dead neurons can be brought back to life by swapping activation functions or implementing synaptic stripping. It is crucial for foundation model training success to proactively monitor neuron health with audits and alerts. In neural networks, some neurons end up outputting near-zero activations across all inputs. These so-called “dead neurons” degrade model capacity because […]
The previous post discussed two ways to compute the nth Fibonacci number. The first is to compute all the Fibonacci numbers up to the nth iteratively using the defining property of Fibonacci numbers Fn + 2 = Fn + Fn + 1 with extended integer arithmetic. The second approach is to use Binet’s formula Fn = round( φn / √ 5 ) where φ is the golden ratio. It’s not clear which approach is more efficient. You could say […]
Detecting anomalies in link streams that represent various kinds of interactions is an important research topic with crucial applications. Because of the lack of ground truth data, proposed methods are mostly evaluated through their ability to detect randomly injected links. In contrast with most proposed methods, that rely on complex approaches raising computational and/or interpretability issues, we show here that trivial graph features and classical learning techniques are sufficient to detect such anomalies extremely well. This basic approach […]