DiffUNet^2 is a bidirectional conditional diffusion model designed for probabilistic scientific spatiotemporal prediction, enabling both forward and backward inference within a single framework. The authors evaluate the model on four datasets covering fluid dynamics, chemical reaction dynamics, and material deformation, demonstrating strong predictive performance and high-quality probabilistic ensembles compared to deterministic and probabilistic baselines. Additionally, DiffUNet^2 supports target‑guided state editing, allowing users to specify and explore states of interest in either temporal direction.
By Mengdi Chu, Jiaxin Yang, Angus G. Forbes, Nathan Debardeleben, Earl Lawrence, Ayan Biswas, Han-Wei Shen
arXiv:2606. 15172v1 Announce Type: new Abstract: Synthesizing realistic time series with generative models has wide-ranging applications in real-world scenarios.
By Zihao Yao, Qi Zheng, Jiankai Zuo, Yaying Zhang
arXiv:2606. 00384v1 Announce Type: new Abstract: Fitting quantitative models to data is a central step in scientific workflows, yet it remains one of the least automated.
By William Rudman, Abhishek Divekar, Kanishk Jain, Sebastian Joseph, Stella S. R. Offner, Matthew Lease, Kyle Mahowald, Greg Durrett, Junyi Jessy Li
arXiv:2605. 29475v2 Announce Type: replace-cross Abstract: Large language models (LLMs) show remarkable potential in scientific hypothesis discovery.
By Hongran An, Zonglin Yang
arXiv:2605. 00972v2 Announce Type: replace-cross Abstract: Earth system science is producing increasingly large, high-dimensional datasets from both physics-based and AI-driven models.
By Nihanth W. Cherukuru, Matt Rehme, Kirsten J. Mayer, David John Gagne, John Schreck, John Clyne, Charlie Becker
Scientific datasets are commonly organized as hierarchical repositories containing heterogeneous and interdependent files, making their inspection, integration, and analysis labor-intensive and reliant on domain expertise. Although large language model (LLM) agents have advanced substantially in planning, reasoning, and tool use, existing research has largely overlooked their ability to interact with real scientific data assets through executable environments.