Dataset And Benchmark Neurips 2025 Model. Poster Presentation at NeurIPS 2023 on a Largescale Dataset of AntigenAntibody Interactions Built on a minimally modified MM-DiT [10] architecture, UniVG seamlessly integrates diverse types of inputs, including text prompts, masks, and existing images, and is able to adapt to different tasks by adjusting its inputs. Our dataset features video clips with a complete set of driving maneuvers, diverse multi-agent interplay, and rich open-world driving knowledge, laying a stepping stone for future world model development.
The State of Data Curation at NeurIPS An Assessment of Dataset Development Practices in the from www.aimodels.fyi
Built on a minimally modified MM-DiT [10] architecture, UniVG seamlessly integrates diverse types of inputs, including text prompts, masks, and existing images, and is able to adapt to different tasks by adjusting its inputs. have been accepted for the Datasets and Benchmarks Track of the NeuRIPS 2025 Conference in Vancouver, BC (CORE Rank A*, Acceptance rate 25%).
The State of Data Curation at NeurIPS An Assessment of Dataset Development Practices in the
We introduce DrivingDojo, the first dataset tailor-made for training interactive world models with complex driving dynamics In this paper, we present UniVG, a diffusion based model that unifies diverse image generation tasks within a single framework Best Paper Award committee for dataset and benchmark track: Yulia Gel, Ludwig.
Two Papers accepted for NeuRIPS 2025 Datasets and Benchmarks Track Universität Mannheim. Built on a minimally modified MM-DiT [10] architecture, UniVG seamlessly integrates diverse types of inputs, including text prompts, masks, and existing images, and is able to adapt to different tasks by adjusting its inputs. A Benchmark for Learning on Temporal Knowledge Graphs and Heterogeneous Graphs" by Julia Gastinger and others
NeurIPS Poster Tenrec A Largescale Multipurpose Benchmark Dataset for Systems. For each self-nomination application for reviewing at NeurIPS 2025, we will consider the following criteria as relevant. have been accepted for the Datasets and Benchmarks Track of the NeuRIPS 2025 Conference in Vancouver, BC (CORE Rank A*, Acceptance rate 25%).