Zobrazeno 1 - 10
of 1 547
pro vyhledávání: '"Huang, James"'
We theoretically investigate the generation of three-photon states with spatial entanglement in cubic nonlinear coupled waveguides using third-order spontaneous parametric down-conversion and quantum walks. Our approach involves independently pumping
Externí odkaz:
http://arxiv.org/abs/2411.07491
Direct preference optimization (DPO) has shown to be an effective method for large language model (LLM) alignment. Recent works have attempted to apply DPO to multimodal scenarios but have found it challenging to achieve consistent improvement. Throu
Externí odkaz:
http://arxiv.org/abs/2406.11839
Autor:
Wang, Fei, Fu, Xingyu, Huang, James Y., Li, Zekun, Liu, Qin, Liu, Xiaogeng, Ma, Mingyu Derek, Xu, Nan, Zhou, Wenxuan, Zhang, Kai, Yan, Tianyi Lorena, Mo, Wenjie Jacky, Liu, Hsiang-Hui, Lu, Pan, Li, Chunyuan, Xiao, Chaowei, Chang, Kai-Wei, Roth, Dan, Zhang, Sheng, Poon, Hoifung, Chen, Muhao
We introduce MuirBench, a comprehensive benchmark that focuses on robust multi-image understanding capabilities of multimodal LLMs. MuirBench consists of 12 diverse multi-image tasks (e.g., scene understanding, ordering) that involve 10 categories of
Externí odkaz:
http://arxiv.org/abs/2406.09411
Autor:
Huang, James Y., Zhou, Wenxuan, Wang, Fei, Morstatter, Fred, Zhang, Sheng, Poon, Hoifung, Chen, Muhao
Despite the strong capabilities of Large Language Models (LLMs) to acquire knowledge from their training corpora, the memorization of sensitive information in the corpora such as copyrighted, harmful, and private content has led to ethical and legal
Externí odkaz:
http://arxiv.org/abs/2404.11045
Autor:
Yan, Tianyi Lorena, Wang, Fei, Huang, James Y., Zhou, Wenxuan, Yin, Fan, Galstyan, Aram, Yin, Wenpeng, Chen, Muhao
Instruction tuning has been used as a promising approach to improve the performance of large language models (LLMs) on unseen tasks. However, current LLMs exhibit limited robustness to unseen instructions, generating inconsistent outputs when the sam
Externí odkaz:
http://arxiv.org/abs/2402.11138
Autor:
Huang, James Y., Sengupta, Sailik, Bonadiman, Daniele, Lai, Yi-an, Gupta, Arshit, Pappas, Nikolaos, Mansour, Saab, Kirchhoff, Katrin, Roth, Dan
Large Language Models (LLMs) are nowadays expected to generate content aligned with human preferences. Current work focuses on alignment at model training time, through techniques such as Reinforcement Learning with Human Feedback (RLHF). However, it
Externí odkaz:
http://arxiv.org/abs/2402.06147
Natural language understanding (NLU) models often suffer from unintended dataset biases. Among bias mitigation methods, ensemble-based debiasing methods, especially product-of-experts (PoE), have stood out for their impressive empirical success. Howe
Externí odkaz:
http://arxiv.org/abs/2305.17627
Traditional sentence embedding models encode sentences into vector representations to capture useful properties such as the semantic similarity between sentences. However, in addition to similarity, sentence semantics can also be interpreted via comp
Externí odkaz:
http://arxiv.org/abs/2305.14599
Parameter-efficient tuning aims at updating only a small subset of parameters when adapting a pretrained model to downstream tasks. In this work, we introduce PASTA, in which we only modify the special token representations (e.g., [SEP] and [CLS] in
Externí odkaz:
http://arxiv.org/abs/2210.04382
Semantic typing aims at classifying tokens or spans of interest in a textual context into semantic categories such as relations, entity types, and event types. The inferred labels of semantic categories meaningfully interpret how machines understand
Externí odkaz:
http://arxiv.org/abs/2205.01826