Papers, modelos e datasets em alta no Hugging Face, além do blog oficial — com leitura editorial em português.
A new benchmark evaluates multimodal large language models' ability to reason over dynamic visual evidence through controlled temporal-logical operations rather than simple object…
DataComp for VLMs (DCVLM) establishes a comprehensive benchmark for evaluating data curation strategies in vision-language models, demonstrating that data mixing rather than filter…
PerceptionRubrics presents a rubric-based evaluation framework that identifies gaps between benchmark scores and real-world performance through atomic auditing and gated scoring me…
A progressive multimodal search agent for knowledge-based visual question answering that adaptively selects search strategies and optimizes through sequence-level reinforcement lea…
Modelo de geração de texto · 397 B de parâmetros — 60.1 mil downloads e 64 curtidas no Hugging Face.
Modelo de geração de texto · 397 B de parâmetros — 94.1 mil downloads e 222 curtidas no Hugging Face.
Modelo de geração de texto · 35 B de parâmetros — 709.3 mil downloads e 379 curtidas no Hugging Face.
Modelo de geração de texto · 9 B de parâmetros — 528.8 mil downloads e 434 curtidas no Hugging Face.
Summer savings are heating up. From the Steam Summer Sale to GeForce NOW membership discounts, this week’s GFN Thursday delivers double the deals and more ways to get the most value from cloud gaming. Plus, Dark Scrolls joins the growing Devolver lineup, alongside Square Enix’s The Adventures of Elliot: The Millennium Tales. They lead the […]
O JoyAI-VL-Interaction é um modelo de 8B que assiste vídeo continuamente, resolve sozinho se deve falar ou ficar em silêncio, e passa problemas difíceis para um modelo secundário — com receita de treinamento, dados e código publicados.
arXiv:2606.25212v1 Announce Type: new Abstract: Accurate physical parameter identification of manipulated objects is fundamental to advanced robotic manipulation and the construction of faithful digital twins. However, acquiring physically consistent inertial and frictional properties from real-world interactions remains challenging due to sensing noise, modeling errors, and limited prior knowledge. This paper presents \textbf{RigPI}, a systematic framework for identifying dynamic parameters of ...
arXiv:2606.25255v1 Announce Type: new Abstract: High-resolution (HR) MRI acquisition is often hampered by scan time constraints, resulting in anisotropic or low-resolution scans (e.g., thick-slice FLAIR) that limit diagnostic accuracy. While deep learning-based super-resolution (SR) methods show promise, they often hallucinate anatomical details, which can compromise brain structural integrity. To mitigate this limitation, we introduce MR-DiffuSR, a Multi-Resolution Diffusion-based Super-Resolut...