arXiv subjects

Fayu Jiang

Publications and source records attributed to Fayu Jiang.

4 recordsLinked to original sources

arXiv2026

Dr.Sai: An agentic AI for real-world physics analysis at BESIII

High Energy Physics (HEP) experiments like BESIII produce petabyte-scale data. Extracting physics results requires complex workflows (simulation, reconstruction, statistical analysis, etc.) that traditionally take experts months or years. Current manual methods are labor-intensive, prone to bias, and limit large-scale systematic scans. As data grows, this paradigm slows discovery. Large Language Models (LLMs) offer a solution. Their natural language understanding and code generation capabilities allow them to interpret scientific tasks and integrate with HEP tools (e.g., ROOT, BOSS) to act as an "AI partner" for autonomous analysis. We present Dr.Sai, an LLM-powered multi-agent system that translates natural language into rigorous physics workflows. As validation, Dr.Sai performed large-scale re-measurements of ten J/psi decay branching fractions - without manual coding. It successfully navigated the real BESIII computing environment and produced results matching established benchmarks. The article details Dr.Sai's architecture, the validation results, and performance evaluation. This work provides a blueprint for autonomous discovery, with relevance to other data-intensive fields like astronomy and genomics.

hep-ex↗

arXiv2015

Interaction between an emerging flux region and a pre-existing fan-spine dome observed by IRIS and SDO

We present multi-wavelength observations of a fan-spine dome in the active region NOAA 11996 with the \textit{Interface Region Imaging Spectrograph} (IRIS) and the Atmospheric Imaging Assembly on board the \textit{Solar Dynamics Observatory} (SDO) on March 9, 2014. The destruction of the fan-spine topology owing to the interaction between its magnetic fields and an nearby emerging flux region (EFR) is firstly observed. The line-of-sight magnetograms from the Helioseismic and Magnetic Imager on board the SDO reveal that the dome is located on the mixed magnetic fields, with its rim rooted in the redundant positive polarity surrounding the minority parasitic negative fields. The fan surface of the dome consists of a filament system and recurring jets are observed along its spine. The jet occurring around 13:54 UT is accompanied with a quasi-circular ribbon that brightens in the clockwise direction along the bottom rim of the dome, which may indicate an occurrence of slipping reconnection in the fan-spine topology. The EFR emerges continuously and meets with the magnetic fields of the dome. Magnetic cancellations take place between the emerging negative polarity and the outer positive polarity of the dome's fields, which lead to the rise of the loop connecting the EFR and brightenings related to the dome. A single Gaussian fit to the profiles of the IRIS SI IV 1394 Å line is used in the analysis. It appears that there are two rising components along the slit, except for the rise in the line-of-sight direction. The cancellation process repeats again and again. Eventually the fan-spine dome is destroyed and a new connectivity is formed. We suggest that magnetic reconnection between the EFR and the magnetic fields of the fan-spine dome in the process is responsible for the destruction of the dome.

astro-ph.SR↗

arXiv2015

Oscillating light wall above a sunspot light bridge

With the high tempo-spatial \emph{Interface Region Imaging Spectrograph} 1330 Å images, we find that many bright structures are rooted in the light bridge of NOAA 12192, forming a \emph{light wall}. The light wall is brighter than the surrounding areas, and the wall top is much brighter than the wall body. The New Vacuum Solar Telescope H$α$ and the \emph{Solar Dynamics Observatory} 171 Å and 131 Å images are also used to study the light wall properties. In 1330 Å, 171 Å, and 131 Å, the top of the wall has a higher emission, while in the H$α$ line, the wall top emission is very low. The wall body corresponds to bright areas in 1330 Å and dark areas in the other lines. The top of the light wall moves upward and downward successively, performing oscillations in height. The deprojected mean height, amplitude, oscillation velocity, and the dominant period are determined to be 3.6 Mm, 0.9 Mm, 15.4 km s$^{-1}$, and 3.9 min, respectively. We interpret the oscillations of the light wall as the leakage of \emph{p}-modes from below the photosphere. The constant brightness enhancement of the wall top implies the existence of some kind of atmospheric heating, e.g., via the persistent small-scale reconnection or the magneto-acoustic waves. In another series of 1330 Å images, we find that the wall top in the upward motion phase is significantly brighter than in the downward phase. This kind of oscillations may be powered by the energy released due to intermittent impulsive magnetic reconnection.

astro-ph.SR↗

arXiv2015

Relationship between EUV microflares and small-scale magnetic fields in the quiet Sun

Microflares are small dynamic signatures observed in X-ray and extreme-ultraviolet channels. Because of their impulsive emission enhancements and wide distribution, they are thought to be closely related to coronal heating. By using the high resolution 171 Å images from the Atmospheric Imaging Assembly and the lines-of-sight magnetograms obtained by the Helioseismic and Magnetic Imager on board the Solar Dynamics Observatory, we trace 10794 microflares in a quiet region near the disk center with a field of view of 960 arcsec $\times$ 1068 arcsec during 24 hr. The microflares have an occurrence rate of 4.4 $\times$ 10$^{3}$ hr$^{-1}$ extrapolated over the whole Sun. Their average brightness, size, and lifetime are 1.7 I$_{0}$(of the quiet Sun), 9.6 Mm$^{2}$, and 3.6 min, respectively. There exists a mutual positive correlation between the microflares' brightness, area and lifetime. In general, the microflares distribute uniformly across the solar disk, but form network patterns locally, which are similar to and matched with the magnetic network structures. Typical cases show that the microflares prefer to occur in magnetic cancellation regions of network boundaries. We roughly calculate the upper limit of energy flux supplied by the microflares and find that the result is still a factor of $\sim$15 below the coronal heating requirement.

astro-ph.SR↗