DGIST’s Electrical Engineering and Computer Science Professor Jang Jae-eun and Professor Kwon Hyuk-jun and their research team have developed a high-efficiency process technology for next-generation AI memory transistors. The work is published online in Advanced Science.

The team developed a nanosecond pulsed laser-based “selective heat treatment method” and “thermal energy minimization control process technology” to overcome the shortcomings of the high-temperature process of ferroelectric field-effect transistors, which have non-volatile memory characteristics, high-speed operation, low power consumption, long lifetime, and durability.

The new technology process enables the realization of heterojunction structures, which are the core technology of next-generation AI semiconductors.

Explore the uncanny valley AI phenomenon, its impact on technology and design, and strategies for creating AI that bridges the gap between human and machine.

Convolutional neural networks (CNNs), with their exceptional image recognition capabilities, have performed outstandingly in the field of AI and notably within platforms like ChatGPT. Recently, a team of Chinese researchers from University of Shanghai for Science and Technology have successfully introduced the concept of CNNs into the field of optics and realized convolutional all-optical neural network, bringing revolutionary progress to AI imaging technology.

Led by Prof. Min Gu and Prof. Qiming Zhang from School of Artificial Intelligence Science and Technology (SAIST) at the University of Shanghai for Science and Technology (USST), the research team has developed an ultrafast convolutional optical neural network (ONN), which achieves efficient and clear imaging of objects behind scattering media without relying on the optical memory effect.

This finding was published in the journal Science Advances, in a paper titled “Memory-less scattering imaging with ultrafast convolutional optical neural networks.”

Numerous electrophysiological experiments have reported that the prefrontal cortex (PFC) is involved in the process of working memory. PFC neurons continue firing to maintain stimulus information in the delay period without external stimuli in working memory tasks. Further findings indicate that while the activity of single neurons exhibits strong temporal and spatial dynamics (heterogeneity), the activity of population neurons can encode spatiotemporal information of stimuli stably and reliably. From the perspective of neural networks, the computational mechanism underlying this phenomenon is not well demonstrated. The main purpose of this paper is to adopt a new strategy to explore the neural computation mechanism of working memory. We used reinforcement learning to train a recurrent neural network model to learn a spatial working memory task.