arXiv:2603.24753v1 Announce Type: new Abstract: Standard vision models treat objects as independent points in Euclidean space, unable to capture hierarchical structure like parts within wholes. We introduce Worldline Slot Attention, which models objects as persistent trajectories through spacetime worldlines, where each object has multiple slots at different hierarchy levels sharing the same spatial position but differing in temporal coordinates. This architecture consistently fails without geometric structure: Euclidean worldlines achieve 0.078 level accuracy, below random chance (0.33), while Lorentzian […]

Ver mais

In CT angiography, the accurate segmentation of abdominal aortic aneurysms (AAAs) is difficult due to large anatomical variability, low-contrast vessel boundaries, and the close proximity of organs whose intensities resemble vascular structures, often leading to false positives. To address these challenges, we propose an anatomy-aware segmentation framework that integrates organ exclusion masks derived from TotalSegmentator into the training process. These masks encode explicit anatomical priors by identifying non-vascular organsand penalizing aneurysm predictions within these regions, thereby guiding the […]

Ver mais

arXiv:2601.10861v1 Announce Type: new Abstract: The transition to electric vehicles (EVs) depends heavily on the reliability of charging infrastructure, yet approximately 1 in 5 drivers report being unable to charge during station visits due to inoperable equipment. While regulatory efforts such as the National Electric Vehicle Infrastructure (NEVI) program have established uptime requirements, these metrics are often simplistic, delayed, and fail to provide the diagnostic granularity needed by Charging Site Operators (CSOs). Despite their pivotal role in maintaining […]

Ver mais

AI companies teaming up with some of the big retail stores is becoming a new trend. For those who still haven’t got why, the reason is pretty obvious: making users shop more smartly. However, the news of Google teaming up with Walmart isn’t just another partnership at all. Both companies are looking to revolutionize how people shop, and this time, it’s not just another AI experiment slapped onto an app. Soon, Gemini will make Walmart shopping easier than […]

Ver mais

arXiv:2601.19908v1 Announce Type: new Abstract: The proliferation of large language models (LLMs) is accelerating the integration of multimodal assistants into edge devices, where inference is executed under stringent latency and energy constraints, often exacerbated by intermittent connectivity. These challenges become particularly acute in the context of multimodal LLMs (MLLMs), as high-dimensional visual inputs are transformed into extensive token sequences, thereby inflating the key-value (KV) cache and imposing substantial data movement overheads to the LLM backbone. To address these […]

Ver mais

arXiv:2603.00621v1 Announce Type: new Abstract: Research in CDCR remains fragmented due to heterogeneous dataset formats, varying annotation standards, and the predominance of the CDCR definition as the event coreference resolution (ECR). To address these challenges, we introduce uCDCR, a unified dataset that consolidates diverse publicly available English CDCR corpora across various domains into a consistent format, which we analyze with standardized metrics and evaluation protocols. uCDCR incorporates both entity and event coreference, corrects known inconsistencies, and enriches datasets […]

Ver mais

Security is treated as a critical priority in modern software organizations. It appears in roadmaps, compliance documents, architectural reviews, and post-incident reports. Yet there is one place where security is still largely invisible: the hiring pipeline. Most engineering teams invest heavily in security tools, audits, and policies, yet devote little effort to evaluating whether new hires can write secure code. The assumption is simple and widespread. Secure coding can be taught later. What matters during hiring is speed, […]

Ver mais