no code implementations • 20 Mar 2024 • Yijian Lu, Aiwei Liu, Dianzhi Yu, Jingjing Li, Irwin King
In this work, we proposed that the influence of token entropy should be fully considered in the watermark detection process, that is, the weight of each token during watermark detection should be adjusted according to its entropy, rather than setting the weights of all tokens to the same value as in previous methods.
no code implementations • 5 Mar 2024 • Yutong Li, Lu Chen, Aiwei Liu, Kai Yu, Lijie Wen
In this work, we firstly focus on the independent literature summarization step and introduce ChatCite, an LLM agent with human workflow guidance for comparative literature summary.
no code implementations • 21 Feb 2024 • Zhiwei He, Binglin Zhou, Hongkun Hao, Aiwei Liu, Xing Wang, Zhaopeng Tu, Zhuosheng Zhang, Rui Wang
Furthermore, we analyze two key factors that contribute to the cross-lingual consistency in text watermarking and propose a defense method that increases the AUC from 0. 67 to 0. 88 under CWRA.
no code implementations • 19 Feb 2024 • Aiwei Liu, Haoping Bai, Zhiyun Lu, Xiang Kong, Simon Wang, Jiulong Shan, Meng Cao, Lijie Wen
In this paper, we propose a method to evaluate the response preference by using the output probabilities of response pairs under contrastive prompt pairs, which could achieve better performance on LLaMA2-7B and LLaMA2-13B compared to RLAIF.
no code implementations • 13 Dec 2023 • Aiwei Liu, Leyi Pan, Yijian Lu, Jingjing Li, Xuming Hu, Xi Zhang, Lijie Wen, Irwin King, Hui Xiong, Philip S. Yu
Text watermarking algorithms play a crucial role in the copyright protection of textual content, yet their capabilities and application scenarios have been limited historically.
1 code implementation • 25 Oct 2023 • Xuming Hu, Junzhe Chen, Aiwei Liu, Shiao Meng, Lijie Wen, Philip S. Yu
Additionally, our method is orthogonal to previous multimodal fusions, and using it on prior SOTA fusions further improves 5. 47% F1.
1 code implementation • 24 Oct 2023 • Shiao Meng, Xuming Hu, Aiwei Liu, Shu'ang Li, Fukun Ma, Yawen Yang, Lijie Wen
However, existing works often struggle to obtain class prototypes with accurate relational semantics: 1) To build prototype for a target relation type, they aggregate the representations of all entity pairs holding that relation, while these entity pairs may also hold other relations, thus disturbing the prototype.
1 code implementation • 10 Oct 2023 • Aiwei Liu, Leyi Pan, Xuming Hu, Shiao Meng, Lijie Wen
In this work, we propose a semantic invariant watermarking method for LLMs that provides both attack robustness and security robustness.
2 code implementations • 30 Jul 2023 • Aiwei Liu, Leyi Pan, Xuming Hu, Shu'ang Li, Lijie Wen, Irwin King, Philip S. Yu
Experiments demonstrate that our algorithm attains high detection accuracy and computational efficiency through neural networks with a minimized number of parameters.
no code implementations • 29 May 2023 • Aiwei Liu, Wei Liu, Xuming Hu, Shuang Li, Fukun Ma, Yawen Yang, Lijie Wen
Based on these observations, we propose a method named \texttt{p-align} to improve the compositional generalization of Text-to-SQL models.
no code implementations • 26 May 2023 • Xuming Hu, Aiwei Liu, Zeqi Tan, Xin Zhang, Chenwei Zhang, Irwin King, Philip S. Yu
These techniques neither preserve the semantic consistency of the original sentences when rule-based augmentations are adopted, nor preserve the syntax structure of sentences when expressing relations using seq2seq models, resulting in less diverse augmentations.
1 code implementation • 22 May 2023 • Shuang Li, Xuming Hu, Aiwei Liu, Yawen Yang, Fukun Ma, Philip S. Yu, Lijie Wen
In this paper, we propose a novel Soft prompt learning framework with the Multilingual Verbalizer (SoftMV) for XNLI.
Cross-Lingual Natural Language Inference Cross-Lingual Transfer
no code implementations • 12 May 2023 • Yawen Yang, Xuming Hu, Fukun Ma, Shu'ang Li, Aiwei Liu, Lijie Wen, Philip S. Yu
Existing works for nested NER ignore the recognition order and boundary position relation of nested entities.
1 code implementation • 12 Mar 2023 • Aiwei Liu, Xuming Hu, Lijie Wen, Philip S. Yu
This paper presents the first comprehensive analysis of ChatGPT's Text-to-SQL ability.
1 code implementation • 31 Oct 2022 • Aiwei Liu, Honghai Yu, Xuming Hu, Shu'ang Li, Li Lin, Fukun Ma, Yawen Yang, Lijie Wen
We propose the first character-level white-box adversarial attack method against transformer models.
no code implementations • 19 Oct 2022 • Xuming Hu, Yong Jiang, Aiwei Liu, Zhongqiang Huang, Pengjun Xie, Fei Huang, Lijie Wen, Philip S. Yu
Data augmentation techniques have been used to alleviate the problem of scarce labeled data in various NER tasks (flat, nested, and discontinuous NER tasks).
1 code implementation • 8 Aug 2022 • Aiwei Liu, Xuming Hu, Li Lin, Lijie Wen
First, we extract a schema linking graph from PLMs through a probing procedure in an unsupervised manner.
1 code implementation • NAACL 2022 • Xuming Hu, Zhijiang Guo, Guanyu Wu, Aiwei Liu, Lijie Wen, Philip S. Yu
The explosion of misinformation spreading in the media ecosystem urges for automated fact-checking.
no code implementations • 31 May 2022 • Shu'ang Li, Xuming Hu, Li Lin, Aiwei Liu, Lijie Wen, Philip S. Yu
Natural Language Inference (NLI) is a growingly essential task in natural language understanding, which requires inferring the relationship between the sentence pairs (premise and hypothesis).