graphics

Large language models, or LLMs, are moving from fast token generation toward deliberate multi-step reasoning. Scaling test-time compute has become a key way to improve performance on complex tasks because it gives models more opportunity to develop intermediate reasoning before producing an answer. However, unconstrained compute scaling frequently leads to a practical failure mode known as &q…

This study aims to develop an interactive Augmented Reality (AR)-based learning medium on chemical bonding material for eleventh-grade students of SMAN 8 Selayar and analyze its effect on students’ motivation and learning outcomes. The study was motivated by the low motivation and achievement in chemistry learning, especially in chemical bonding material, which is abstract and difficult to unders…

Abstract One-stream transformer trackers have received widespread attention for their excellent discriminatory ability. However, most of the existing trackers try to mine more information about the target while ignoring the exploitation of the background around it. In this work We propose a single-stream progressive background elimination transformer for target tracking. This model employs a prog…

International audience

We present NPC Nano, a 501M-parameter decoder-only language model pretrained from random initialization on 8.93B tokens using a single NVIDIA A40 GPU. We document the pretraining recipe, a label-shift bug encountered during training and the pre-launch sanity gate that prevents its recurrence, an identity layer methodology with empirically recalibrated capability gates, and a four-experiment chara…

Abstract Detection of deepfakes has become a more difficult task due to the Escalating Sophistication of Reproductive Reproductions, particularly D-F architectures, such as the existing methods, which have problems with cross-dataset generalization because they rely on single-stream deep features and naive concatenation approaches. In this Paper, we present AFFD-Net (Attention-Guided Feature Fusi…

This paper presents a pilot study on the automated recognition of historical bookbinding tools using deep learning and synthetic image generation. The work focuses on the documented collection of Czech binder Jenda Rajman (1892–1965), whose complete set of metal stamping tools provides an ideal reference dataset. Each tool leaves a characteristic blind-stamped impression on leather bindings, form…

Abstract Videogrammetry can quantify head acceleration events in sport, but because standard datasets lack the large rotations, rapid motion, and frequent occlusion characteristic of sports collisions, the accuracy of modern deep learning pose estimators in this context remains unclear. This study addresses this gap by benchmarking three models for monocular head pose estimation during controlled…

Smart warehouses rely on fleets of autonomous mobile robots that must continually assign tasks, plan paths, avoid collisions, and maintain battery energy. Existing lifelong multi-agent path finding studies often emphasize travel cost or makespan, while practical deployments also involve charging, payload-dependent energy use, turning and waiting costs, and congestion. This paper presents an energ…

Genesis as Algorithm is a Python-based MIDI composition project organized into 74 Thematic Units. The work combines numerical matrices, te'amim-inspired melodic modules, and a four-channel polyphonic architecture to generate a structured algorithmic musical output. It is intended as an artistic and research-oriented software artifact for archival publication and reproducible use.

Hardware tessellation as we know it today (Dx11-style) had its origins on the Xbox 360, which released in 2005. Time flies, right? It was a natural step in the evolution toward film-quality realtime rendering. After all, tessellation was a key component of the original Pixar Reyes paper [7]. Now that it’s 20+ years later, we have more experience with the algorithm, and hardware tessellation has n…

This study proposes an innovative approach to detecting structural matches in programming codes, which addresses the fundamental limitation of existing methods – their sensitivity to syntactic changes while maintaining logical equivalence. A hybrid architecture integrating semantic normalization through large language models (LLMs) with multispecies graph representation (AST, CFG, DFG) and embedd…