Zobrazeno 1 - 10
of 808
pro vyhledávání: '"Shah,Ankit A"'
Autor:
Wang, Yaxuan, Wei, Jiaheng, Liu, Chris Yuhao, Pang, Jinlong, Liu, Quan, Shah, Ankit Parag, Bao, Yujia, Liu, Yang, Wei, Wei
Unlearning in Large Language Models (LLMs) is essential for ensuring ethical and responsible AI use, especially in addressing privacy leak, bias, safety, and evolving regulations. Existing approaches to LLM unlearning often rely on retain data or a r
Externí odkaz:
http://arxiv.org/abs/2410.11143
Autor:
Pang, Jinlong, Wei, Jiaheng, Shah, Ankit Parag, Zhu, Zhaowei, Wang, Yaxuan, Qian, Chen, Liu, Yang, Bao, Yujia, Wei, Wei
Instruction tuning is critical for adapting large language models (LLMs) to downstream tasks, and recent studies have demonstrated that small amounts of human-curated data can outperform larger datasets, challenging traditional data scaling laws. Whi
Externí odkaz:
http://arxiv.org/abs/2410.10877
Autor:
Raghavan, Ksheeraja, Gode, Samiran, Shah, Ankit, Raghavan, Surabhi, Burgard, Wolfram, Raj, Bhiksha, Singh, Rita
We introduce a novel, general-purpose audio generation framework specifically designed for anomaly detection and localization. Unlike existing datasets that predominantly focus on industrial and machine-related sounds, our framework focuses a broader
Externí odkaz:
http://arxiv.org/abs/2410.03904
Autor:
Liu, Minghao, Di, Zonglin, Wei, Jiaheng, Wang, Zhongruo, Zhang, Hengxiang, Xiao, Ruixuan, Wang, Haoyu, Pang, Jinlong, Chen, Hao, Shah, Ankit, Wei, Hongxin, He, Xinlei, Zhao, Zhaowei, Wang, Haobo, Feng, Lei, Wang, Jindong, Davis, James, Liu, Yang
Large-scale data collection is essential for developing personalized training data, mitigating the shortage of training data, and fine-tuning specialized models. However, creating high-quality datasets quickly and accurately remains a challenge due t
Externí odkaz:
http://arxiv.org/abs/2408.11338
Autor:
Bao, Yujia, Shah, Ankit Parag, Narang, Neeru, Rivers, Jonathan, Maksey, Rajeev, Guan, Lan, Barrere, Louise N., Evenson, Shelley, Basole, Rahul, Miao, Connie, Mehta, Ankit, Boulay, Fabien, Park, Su Min, Pearson, Natalie E., Joy, Eldhose, He, Tiger, Thakur, Sumiran, Ghosal, Koustav, On, Josh, Morrison, Phoebe, Major, Tim, Wang, Eva Siqi, Escobar, Gina, Wei, Jiaheng, Weerasooriya, Tharindu Cyril, Song, Queena, Lashkevich, Daria, Chen, Clare, Kim, Gyuhak, Yin, Dengpan, Hejna, Don, Nomeli, Mo, Wei, Wei
This paper introduces Fortune Analytics Language Model (FALM). FALM empowers users with direct access to comprehensive business analysis, including market trends, company performance metrics, and expert insights. Unlike generic LLMs, FALM leverages a
Externí odkaz:
http://arxiv.org/abs/2406.06559
Audio-visual automatic speech recognition (AV-ASR) models are very effective at reducing word error rates on noisy speech, but require large amounts of transcribed AV training data. Recently, audio-visual self-supervised learning (SSL) approaches hav
Externí odkaz:
http://arxiv.org/abs/2312.09369
Human immunodeficiency virus (HIV) is a major public health concern in the United States, with about 1.2 million people living with HIV and 35,000 newly infected each year. There are considerable geographical disparities in HIV burden and care access
Externí odkaz:
http://arxiv.org/abs/2311.00855
Autor:
Konan, Joseph, Agnihotri, Shikhar, Bhargave, Ojas, Han, Shuo, Zeng, Yunyang, Shah, Ankit, Raj, Bhiksha
Within the ambit of VoIP (Voice over Internet Protocol) telecommunications, the complexities introduced by acoustic transformations merit rigorous analysis. This research, rooted in the exploration of proprietary sender-side denoising effects, meticu
Externí odkaz:
http://arxiv.org/abs/2310.07161
Autor:
Shah, Muhammad Ahmed, Sharma, Roshan, Dhamyal, Hira, Olivier, Raphael, Shah, Ankit, Konan, Joseph, Alharthi, Dareen, Bukhari, Hazim T, Baali, Massa, Deshmukh, Soham, Kuhlmann, Michael, Raj, Bhiksha, Singh, Rita
It has been shown that Large Language Model (LLM) alignments can be circumvented by appending specially crafted attack suffixes with harmful queries to elicit harmful responses. To conduct attacks against private target models whose characterization
Externí odkaz:
http://arxiv.org/abs/2310.04445
Autor:
Chen, Hao, Wang, Jindong, Shah, Ankit, Tao, Ran, Wei, Hongxin, Xie, Xing, Sugiyama, Masashi, Raj, Bhiksha
Pre-training on large-scale datasets and then fine-tuning on downstream tasks have become a standard practice in deep learning. However, pre-training data often contain label noise that may adversely affect the generalization of the model. This paper
Externí odkaz:
http://arxiv.org/abs/2309.17002