In a paper published in National Science Review, a research team from Institute of Automation, Chinese Academy of Sciences and Nanjing University present an overview of the historical developments in Generative Artificial Intelligence (Generative AI). They grouped the developments of Generative AI into four categories: 1) rule-based generative systems, 2) model-based generative algorithms, 3) deep generative methodologies, and 4) foundation models. They also described potential research directions aimed at better utilizing, understanding, and harnessing Generative AI technologies.

Transfer learning, as a new machine learning methodology, may solve problems in related but different domains by using existing knowledge, and it is often applied to transfer training data from another domain for model training in the case of insufficient training data. In recent years, an increasing number of researchers who engage in brain-computer interface (BCI), have focused on using transfer learning to make most of the available electroencephalogram data from different subjects, effectively reducing the cost of expensive data acquisition and labeling as well as greatly improving the learning performance of the model. This paper surveys the development of transfer learning and reviews the transfer learning approaches in BCI. In addition, according to the “what to transfer” question in transfer learning, this review is organized into three contexts: instance-based transfer learning, parameter-based transfer learning, and feature-based transfer learning. Furthermore, the current transfer learning applications in BCI research are summarized in terms of the transfer learning methods, datasets, evaluation performance, etc. At the end of the paper, the questions to be solved in future research are put forward, laying the foundation for the popularization and in-depth research of transfer learning in BCI.

Understanding photogenerated carrier transport in 2D perovskites, especially surface states, is challenging with conventional time-resolved techniques. Scientists at KAUST utilized scanning ultrafast electron microscopy (SUEM) with groundbreaking surface sensitivity to disclose carrier diffusion rates of ~30 cm²/s for n=1, 180 cm²/s for n=2, and 470 cm²/s for n=3, which are notably higher than bulk. This highlights the SUEM’s potential for advancing the understanding of carrier dynamics. Density Functional Theory (DFT) confirms broader carrier transmission channels at the surface, offering key insights for optimizing 2D perovskite optoelectronic devices.

In a paper published in National Science Review, a research team from the Chinese Academy of Sciences investigated impact melt rocks from the Chang'e-6 lunar soils and determined that the Moon's South Pole-Aitken basin formed 4.25 billion years ago. China’s Chang'e-6 mission, launched on May 3, 2024, landed on June 2, and returned on June 25, collecting 1935.3 grams of lunar soil samples – the first ever returned from the Moon's far side.

Inspired by the concept of harmonious coexistence between humans and nature, a new concept, biogenic construction for civil engineering is proposed in this paper, as well as the definition, framework and four components.

Researchers at Saratov State University and their colleagues have developed an innovative approach to improving the visualization of subsurface blood vessels using laser speckle contrast imaging (LSCI) enhanced with principal component analysis (PCA).

This study explores the effects of Alpiniae oxyphyllae Fructus (AOF) on renal lipid deposition in diabetic kidney disease and elucidate its molecular mechanisms.

Research team proposed a UAV system for SAR missions based on consortium blockchain with sharding. It improves adaptability by dynamically creating shards and enhances interoperability via Merkle proof-based cross-shard interaction.