Search

Your search keyword '"Li, Yaxi"' showing total 427 results
Start Over Author "Li, Yaxi"
427 results on '"Li, Yaxi"'

1. SLM Meets LLM: Balancing Latency, Interpretability and Consistency in Hallucination Detection

Author

Hu, Mengya, Xu, Rui, Lei, Deren, Li, Yaxi, Wang, Mingyu, Ching, Emily, Kamal, Eslam, and Deng, Alex

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract

Large language models (LLMs) are highly capable but face latency challenges in real-time applications, such as conducting online hallucination detection. To overcome this issue, we propose a novel framework that leverages a small language model (SLM) classifier for initial detection, followed by a LLM as constrained reasoner to generate detailed explanations for detected hallucinated content. This study optimizes the real-time interpretable hallucination detection by introducing effective prompting techniques that align LLM-generated explanations with SLM decisions. Empirical experiment results demonstrate its effectiveness, thereby enhancing the overall user experience., Comment: preprint under review

Published
2024

4. Chain of Natural Language Inference for Reducing Large Language Model Ungrounded Hallucinations

Author

Lei, Deren, Li, Yaxi, Hu, Mengya, Wang, Mingyu, Yun, Vincent, Ching, Emily, and Kamal, Eslam

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

Large language models (LLMs) can generate fluent natural language texts when given relevant documents as background context. This ability has attracted considerable interest in developing industry applications of LLMs. However, LLMs are prone to generate hallucinations that are not supported by the provided sources. In this paper, we propose a hierarchical framework to detect and mitigate such ungrounded hallucination. Our framework uses Chain of Natural Language Inference (CoNLI) for hallucination detection and hallucination reduction via post-editing. Our approach achieves state-of-the-art performance on hallucination detection and enhances text quality through rewrite, using LLMs without any fine-tuning or domain-specific prompt engineering. We show that this simple plug-and-play framework can serve as an effective choice for hallucination detection and reduction, achieving competitive performance across various contexts., Comment: The source code is available at https://github.com/microsoft/CoNLI_hallucination

Published
2023

19. Faslpr gene dosage tunes the extent of lymphoproliferation and T cell differentiation in lupus

Author

Bohat, Ritu, Liang, Xiaofang, Chen, Yanping, Xu, Chunyu, Zheng, Ningbo, Guerrero, Ashley, Hou, Jiakai, Jaffery, Roshni, Egan, Nicholas A., Li, Yaxi, Tang, Yitao, Unsal, Esra, Robles, Adolfo, Chen, Si, Major, Angela M., Elldakli, Hadil, Chung, Sang-Hyuk, Liang, Han, Hicks, M. John, Du, Yong, Lin, Jamie S., Chen, Xiqun, Mohan, Chandra, and Peng, Weiyi

Published
2024
Full Text
View/download PDF

41. Study of several probability distribution functions for the Klein–Kramers equation.

Author

Li, Yaxi and Kai, Yue

Subjects
*WEIBULL distribution, *POWER law (Mathematics), *GAUSSIAN distribution, *EQUATIONS, *GAMMA distributions, *SEPARATION of variables, *DISTRIBUTION (Probability theory), *PROBABILITY density function
Abstract

In this paper, we take variable separation method to study Klein–Kramers (KK) equation. By choosing different eigenvalues and noise functions, we can get different probability density functions (PDFs) of KK equation. These PDFs contain not only normal distributions but also other distributions that correspond to anomalous diffusion phenomena. For example, power-law distribution, truncated Cauchy–Lorentz distribution, Weibull distribution, log-logistic distribution, Gamma distribution. We also show the 3D and 2D profiles of these PDFs to analyze the corresponding dynamic properties and illustrate the possible practical applications of these results. In addition, we also find some exact solutions that are not PDFs. They are also listed to ensure the completeness of the results and to illustrate the potential applications of these exact solutions. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

42. Modification of the CuO electronic structure for enhanced selective electrochemical CO2 reduction to ethylene.

Author

Wu, Xin, Tong, Zhuang, Liu, Yunliang, Li, Yaxi, Cheng, Yuanyuan, Yu, Jingwen, Cao, Peng, Zhuang, Chunqiang, Shi, Qiuzhong, Liu, Naiyun, Liu, Xiang, Liang, Hongyu, and Li, Haitao

Abstract

Electrochemical carbon dioxide reduction reaction (CO2RR) can produce value-added hydrocarbons from renewable electricity, providing a sustainable and promising approach to meet dual-carbon targets and alleviate the energy crisis. However, it is still challenging to improve the selectivity and stability of the products, especially the C2+ products. Here we propose to modulate the electronic structure of copper oxide (CuO) through lattice strain construction by zinc (Zn) doping to improve the selectivity of the catalyst to ethylene. Combined performance and in situ characterization analyses show that the compressive strain generated within the CuO lattice and the electronic structure modulation by Zn doping enhances the adsorption of the key intermediate ⋆CO, thereby increasing the intrinsic activity of CO2RR and inhibiting the hydrogen precipitation reaction. Among the best catalysts had significantly improved ethylene selectivity of 60.5% and partial current density of 500 mA·cm−2, and the highest C2+ Faraday efficiency of 71.47%. This paper provides a simple idea to study the modulation of CO2RR properties by heteroatom doped and lattice strain. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results