Zobrazeno 1 - 10
of 87
pro vyhledávání: '"Dai, Zhongxiang"'
Contextual dueling bandit is used to model the bandit problems, where a learner's goal is to find the best arm for a given context using observed noisy preference feedback over the selected arms for the past contexts. However, existing algorithms ass
Externí odkaz:
http://arxiv.org/abs/2407.17112
Autor:
Xu, Xinyi, Wu, Zhaoxuan, Qiao, Rui, Verma, Arun, Shu, Yao, Wang, Jingtan, Niu, Xinyuan, He, Zhenfeng, Chen, Jiangwei, Zhou, Zijian, Lau, Gregory Kang Ruey, Dao, Hieu, Agussurja, Lucas, Sim, Rachael Hwee Ling, Lin, Xiaoqiang, Hu, Wenyang, Dai, Zhongxiang, Koh, Pang Wei, Low, Bryan Kian Hsiang
This position paper proposes a data-centric viewpoint of AI research, focusing on large language models (LLMs). We start by making the key observation that data is instrumental in the developmental (e.g., pretraining and fine-tuning) and inferential
Externí odkaz:
http://arxiv.org/abs/2406.14473
Autor:
Lin, Xiaoqiang, Dai, Zhongxiang, Verma, Arun, Ng, See-Kiong, Jaillet, Patrick, Low, Bryan Kian Hsiang
Large language models (LLMs) have demonstrated remarkable performances in various tasks. However, the performance of LLMs heavily depends on the input prompt, which has given rise to a number of recent works on prompt optimization. However, previous
Externí odkaz:
http://arxiv.org/abs/2405.17346
Autor:
Wu, Zhaoxuan, Lin, Xiaoqiang, Dai, Zhongxiang, Hu, Wenyang, Shu, Yao, Ng, See-Kiong, Jaillet, Patrick, Low, Bryan Kian Hsiang
Large language models (LLMs) have shown impressive capabilities in real-world applications. The capability of in-context learning (ICL) allows us to adapt an LLM to downstream tasks by including input-label exemplars in the prompt without model fine-
Externí odkaz:
http://arxiv.org/abs/2405.16122
Neural architecture search (NAS) has become a key component of AutoML and a standard tool to automate the design of deep neural networks. Recently, training-free NAS as an emerging paradigm has successfully reduced the search costs of standard traini
Externí odkaz:
http://arxiv.org/abs/2403.07591
Autor:
Hu, Wenyang, Shu, Yao, Yu, Zongmin, Wu, Zhaoxuan, Lin, Xiangqiang, Dai, Zhongxiang, Ng, See-Kiong, Low, Bryan Kian Hsiang
The efficacy of large language models (LLMs) in understanding and generating natural language has aroused a wide interest in developing prompt-based methods to harness the power of black-box LLMs. Existing methodologies usually prioritize a global op
Externí odkaz:
http://arxiv.org/abs/2403.02993
We study a novel variant of the parameterized bandits problem in which the learner can observe additional auxiliary feedback that is correlated with the observed reward. The auxiliary feedback is readily available in many real-life applications, e.g.
Externí odkaz:
http://arxiv.org/abs/2311.02715
Autor:
Dai, Zhongxiang, Nguyen, Quoc Phong, Tay, Sebastian Shenghong, Urano, Daisuke, Leong, Richalynn, Low, Bryan Kian Hsiang, Jaillet, Patrick
Many real-world experimental design problems (a) evaluate multiple experimental conditions in parallel and (b) replicate each condition multiple times due to large and heteroscedastic observation noise. Given a fixed total budget, this naturally indu
Externí odkaz:
http://arxiv.org/abs/2311.01195
Autor:
Dai, Zhongxiang, Lau, Gregory Kang Ruey, Verma, Arun, Shu, Yao, Low, Bryan Kian Hsiang, Jaillet, Patrick
Kernelized bandits, also known as Bayesian optimization (BO), has been a prevalent method for optimizing complicated black-box reward functions. Various BO algorithms have been theoretically shown to enjoy upper bounds on their cumulative regret whic
Externí odkaz:
http://arxiv.org/abs/2310.05373
Autor:
Wang, Jingtan, Lu, Xinyang, Zhao, Zitong, Dai, Zhongxiang, Foo, Chuan-Sheng, Ng, See-Kiong, Low, Bryan Kian Hsiang
The impressive performances of large language models (LLMs) and their immense potential for commercialization have given rise to serious concerns over the intellectual property (IP) of their training data. In particular, the synthetic texts generated
Externí odkaz:
http://arxiv.org/abs/2310.00646