arXiv subjects

Siyuan Wu

Publications and source records attributed to Siyuan Wu.

22 records · Page 2Linked to original sources

arXiv2024

I Think, Therefore I am: Benchmarking Awareness of Large Language Models Using AwareBench

Do large language models (LLMs) exhibit any forms of awareness similar to humans? In this paper, we introduce AwareBench, a benchmark designed to evaluate awareness in LLMs. Drawing from theories in psychology and philosophy, we define awareness in LLMs as the ability to understand themselves as AI models and to exhibit social intelligence. Subsequently, we categorize awareness in LLMs into five dimensions, including capability, mission, emotion, culture, and perspective. Based on this taxonomy, we create a dataset called AwareEval, which contains binary, multiple-choice, and open-ended questions to assess LLMs' understandings of specific awareness dimensions. Our experiments, conducted on 13 LLMs, reveal that the majority of them struggle to fully recognize their capabilities and missions while demonstrating decent social intelligence. We conclude by connecting awareness of LLMs with AI alignment and safety, emphasizing its significance to the trustworthy and ethical development of LLMs. Our dataset and code are available at https://github.com/HowieHwong/Awareness-in-LLM.

cs.CL↗

arXiv2023

Excitation of extraordinary modes inside the source of Saturn's kilometric radiation

The electron cyclotron maser instability (ECMI) of extraordinary mode waves was investigated with the parameters observed in Saturn's kilometric radiation (SKR) sources. Previous studies employed simplified dispersion relations, and did not consider the excitation of the relativistic (R) mode. This mode is introduced by considering the relativistic effect in plasmas consisting of both cold and hot electrons. Using particle-in-cell simulations, we investigated the excitation of R and X modes based on the measured data. Using the reported value of the density ratio of energetic to total electrons $n_e/n_0=24\%$, the most unstable mode is the R mode. The escaping X-mode emissions are amplified only if the energetic electrons are dominant with $n_e/n_0 \ge 90\%$. For these cases, only the X mode is excited and the R mode disappears due to its strong coupling. The results are well in line with the linear kinetic theory of ECMI. The properties of both the R and X modes are consistent with the observed SKR emissions. This raises questions about the nature of the measured electric field fluctuations within ``presumed'' SKR sources. The study provides new insights into the ECMI process relevant to SKR emission mechanisms.

astro-ph.EP↗

arXiv2023

LiY(SO$_4$)$_2$: A Superionic Material Synthesized by Superionic State Hidden in no-Superionic Literature

A potential superionic material LiY(SO$_4$)$_2$ has been excavated from the published literatures because its synthesis method and experiment data implied it exists the superionic state. We use \textit{ab initio} calculation to analyzing the differences between solid state and superionic state. We found the diffusion of Li$^+$ from the lattice site to the interstitial site will change the nearest neighbor numbers of O atom from 4 to 8. In order to reduce energy, the reorientation of SO$_4^{2-}$ must exist accompany with the diffusion of Li$^+$ so the nearest neighbor number of O will keep about 5 in the superionic state. Our work not only presents an example for discovering materials from literatures based on prior knowledge but also reveals the micromechanism of cation-anion coupled dynamics for superionic state.

cond-mat.mtrl-sci↗

arXiv2021

FAST-Dynamic-Vision: Detection and Tracking Dynamic Objects with Event and Depth Sensing

The development of aerial autonomy has enabled aerial robots to fly agilely in complex environments. However, dodging fast-moving objects in flight remains a challenge, limiting the further application of unmanned aerial vehicles (UAVs). The bottleneck of solving this problem is the accurate perception of rapid dynamic objects. Recently, event cameras have shown great potential in solving this problem. This paper presents a complete perception system including ego-motion compensation, object detection, and trajectory prediction for fast-moving dynamic objects with low latency and high precision. Firstly, we propose an accurate ego-motion compensation algorithm by considering both rotational and translational motion for more robust object detection. Then, for dynamic object detection, an event camera-based efficient regression algorithm is designed. Finally, we propose an optimizationbased approach that asynchronously fuses event and depth cameras for trajectory prediction. Extensive real-world experiments and benchmarks are performed to validate our framework. Moreover, our code will be released to benefit related researches.

cs.RO↗