Zobrazeno 1 - 9
of 9
pro vyhledávání: '"Kamal, Eslam"'
Autor:
Hu, Mengya, Xu, Rui, Lei, Deren, Li, Yaxi, Wang, Mingyu, Ching, Emily, Kamal, Eslam, Deng, Alex
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)
Externí odkaz:
http://arxiv.org/abs/2408.12748
Autor:
Magooda, Ahmed, Helyar, Alec, Jackson, Kyle, Sullivan, David, Atalla, Chad, Sheng, Emily, Vann, Dan, Edgar, Richard, Palangi, Hamid, Lutz, Roman, Kong, Hongliang, Yun, Vincent, Kamal, Eslam, Zarfati, Federico, Wallach, Hanna, Bird, Sarah, Chen, Mei
We present a framework for the automated measurement of responsible AI (RAI) metrics for large language models (LLMs) and associated products and services. Our framework for automatically measuring harms from LLMs builds on existing technical and soc
Externí odkaz:
http://arxiv.org/abs/2310.17750
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 gene
Externí odkaz:
http://arxiv.org/abs/2310.03951
Autor:
Chan, Aaron, Kharkar, Anant, Moghaddam, Roshanak Zilouchian, Mohylevskyy, Yevhen, Helyar, Alec, Kamal, Eslam, Elkamhawy, Mohamed, Sundaresan, Neel
Software vulnerabilities bear enterprises significant costs. Despite extensive efforts in research and development of software vulnerability detection methods, uncaught vulnerabilities continue to put software owners and users at risk. Many current v
Externí odkaz:
http://arxiv.org/abs/2306.01754
Autor:
Shukla, Swadheen, Liden, Lars, Shayandeh, Shahin, Kamal, Eslam, Li, Jinchao, Mazzola, Matt, Park, Thomas, Peng, Baolin, Gao, Jianfeng
Traditionally, industry solutions for building a task-oriented dialog system have relied on helping dialog authors define rule-based dialog managers, represented as dialog flows. While dialog flows are intuitively interpretable and good for simple sc
Externí odkaz:
http://arxiv.org/abs/2004.04305
Publikováno v:
In Alexandria Engineering Journal December 2019 58(4):1257-1264
Publikováno v:
Computational Linguistics & Intelligent Text Processing 16th International Conference, CICLing 2015, Cairo, Egypt, April 14-20, 2015, Proceedings, Part I; 2015, p98-111, 14p
Publikováno v:
Artificial Neural Networks in Pattern Recognition: 6th IAPR TC 3 International Workshop, ANNPR 2014, Montreal, QC, Canada, October 6-8, 2014. Proceedings; 2014, p23-34, 12p
Publikováno v:
Natural Language Processing & Information Systems (9783642388231); 2013, p53-64, 12p