no code implementations • ACL 2022 • Changzai Pan, Maosong Sun, Ke Deng
Processing open-domain Chinese texts has been a critical bottleneck in computational linguistics for decades, partially because text segmentation and word discovery often entangle with each other in this challenging scenario.
1 code implementation • ACL 2022 • Xu Han, Guoyang Zeng, Weilin Zhao, Zhiyuan Liu, Zhengyan Zhang, Jie zhou, Jun Zhang, Jia Chao, Maosong Sun
In recent years, large-scale pre-trained language models (PLMs) containing billions of parameters have achieved promising results on various NLP tasks.
no code implementations • EMNLP 2021 • Mieradilijiang Maimaiti, Yang Liu, Yuanhang Zheng, Gang Chen, Kaiyu Huang, Ji Zhang, Huanbo Luan, Maosong Sun
Besides, the robustness of the previous neural methods is limited by the large-scale annotated data.
no code implementations • EMNLP 2021 • Yuanhang Zheng, Zhixing Tan, Meng Zhang, Mieradilijiang Maimaiti, Huanbo Luan, Maosong Sun, Qun Liu, Yang Liu
Quality estimation (QE) of machine translation (MT) aims to evaluate the quality of machine-translated sentences without references and is important in practical applications of MT.
1 code implementation • EMNLP 2021 • Yuan YAO, Jiaju Du, Yankai Lin, Peng Li, Zhiyuan Liu, Jie zhou, Maosong Sun
Existing relation extraction (RE) methods typically focus on extracting relational facts between entity pairs within single sentences or documents.
1 code implementation • Findings (ACL) 2022 • Yining Ye, Fanchao Qi, Zhiyuan Liu, Maosong Sun
However, all existing sememe prediction studies ignore the hierarchical structures of sememes, which are important in the sememe-based semantic description system.
1 code implementation • ACL 2022 • Biru Zhu, Yujia Qin, Fanchao Qi, Yangdong Deng, Zhiyuan Liu, Maosong Sun, Ming Gu
To validate our viewpoints, we design two methods to evaluate the robustness of FMS: (1) model disguise attack, which post-trains an inferior PTM with a contrastive objective, and (2) evaluation data selection, which selects a subset of the data points for FMS evaluation based on K-means clustering.
no code implementations • Findings (EMNLP) 2021 • Bo Ouyang, Wenbing Huang, Runfa Chen, Zhixing Tan, Yang Liu, Maosong Sun, Jihong Zhu
Knowledge representation learning (KRL) has been used in plenty of knowledge-driven tasks.
1 code implementation • ACL 2022 • Xu Han, Yuqi Luo, Weize Chen, Zhiyuan Liu, Maosong Sun, Zhou Botong, Hao Fei, Suncong Zheng
In this paper, we propose a cross-lingual contrastive learning framework to learn FGET models for low-resource languages.
1 code implementation • 7 May 2024 • Chen Qian, Jiahao Li, Yufan Dang, Wei Liu, Yifei Wang, Zihao Xie, Weize Chen, Cheng Yang, Yingli Zhang, Zhiyuan Liu, Maosong Sun
We propose two fundamental patterns: the successive pattern, refining based on nearest experiences within a task batch, and the cumulative pattern, acquiring experiences across all previous task batches.
no code implementations • 28 Apr 2024 • Zhili Cheng, Zhitong Wang, Jinyi Hu, Shengding Hu, An Liu, Yuge Tu, Pengkai Li, Lei Shi, Zhiyuan Liu, Maosong Sun
Despite advancements in Large Language Models (LLMs) and Large Multimodal Models (LMMs), their integration into language-grounded, human-like embodied agents remains incomplete, hindering complex real-life task performance in physical environments.
no code implementations • 19 Apr 2024 • Pablo Biedma, Xiaoyuan Yi, Linus Huang, Maosong Sun, Xing Xie
Recent advancements in Large Language Models (LLMs) have revolutionized the AI field but also pose potential safety and ethical risks.
1 code implementation • 11 Apr 2024 • Chaoqun He, Renjie Luo, Shengding Hu, Yuanqian Zhao, Jie zhou, Hanghao Wu, Jiajie Zhang, Xu Han, Zhiyuan Liu, Maosong Sun
The rapid development of LLMs calls for a lightweight and easy-to-use framework for swift evaluation deployment.
2 code implementations • 9 Apr 2024 • Shengding Hu, Yuge Tu, Xu Han, Chaoqun He, Ganqu Cui, Xiang Long, Zhi Zheng, Yewei Fang, Yuxiang Huang, Weilin Zhao, Xinrong Zhang, Zheng Leng Thai, Kaihuo Zhang, Chongyi Wang, Yuan YAO, Chenyang Zhao, Jie zhou, Jie Cai, Zhongwu Zhai, Ning Ding, Chao Jia, Guoyang Zeng, Dahai Li, Zhiyuan Liu, Maosong Sun
For data scaling, we introduce a Warmup-Stable-Decay (WSD) learning rate scheduler (LRS), conducive to continuous training and domain adaptation.
no code implementations • 3 Apr 2024 • Zhiyuan Wen, Jiannong Cao, Jiaxing Shen, Ruosong Yang, Shuaiqi Liu, Maosong Sun
Therefore, we propose a new task, Personality-affected Emotion Generation, to generate emotion based on the personality given to the dialog system and further investigate a solution through the personality-affected mood transition.
1 code implementation • 2 Apr 2024 • Lifan Yuan, Ganqu Cui, Hanbin Wang, Ning Ding, Xingyao Wang, Jia Deng, Boji Shan, Huimin Chen, Ruobing Xie, Yankai Lin, Zhenghao Liu, BoWen Zhou, Hao Peng, Zhiyuan Liu, Maosong Sun
We introduce Eurus, a suite of large language models (LLMs) optimized for reasoning.
1 code implementation • 26 Mar 2024 • Yingfa Chen, Zhengyan Zhang, Xu Han, Chaojun Xiao, Zhiyuan Liu, Chen Chen, Kuai Li, Tao Yang, Maosong Sun
Large language models (LLMs) can make predictions using parametric knowledge--knowledge encoded in the model weights--or contextual knowledge--knowledge presented in the context.
2 code implementations • 18 Mar 2024 • Ruyi Xu, Yuan YAO, Zonghao Guo, Junbo Cui, Zanlin Ni, Chunjiang Ge, Tat-Seng Chua, Zhiyuan Liu, Maosong Sun, Gao Huang
To address the challenges, we present LLaVA-UHD, a large multimodal model that can efficiently perceive images in any aspect ratio and high resolution.
1 code implementation • 14 Mar 2024 • Sun Ao, Weilin Zhao, Xu Han, Cheng Yang, Zhiyuan Liu, Chuan Shi, Maosong Sun, Shengnan Wang, Teng Su
Effective attention modules have played a crucial role in the success of Transformer-based large language models (LLMs), but the quadratic time and memory complexities of these attention modules also pose a challenge when processing long sequences.
no code implementations • 13 Mar 2024 • Ning Ding, Yulin Chen, Ganqu Cui, Xingtai Lv, Weilin Zhao, Ruobing Xie, BoWen Zhou, Zhiyuan Liu, Maosong Sun
Underlying data distributions of natural language, programming code, and mathematical symbols vary vastly, presenting a complex challenge for large language models (LLMs) that strive to achieve high performance across all three domains simultaneously.
2 code implementations • 12 Mar 2024 • Zhicheng Guo, Sijie Cheng, Hao Wang, Shihao Liang, Yujia Qin, Peng Li, Zhiyuan Liu, Maosong Sun, Yang Liu
The virtual API server contains a caching system and API simulators which are complementary to alleviate the change in API status.
no code implementations • 7 Mar 2024 • Xinpeng Wang, Shitong Duan, Xiaoyuan Yi, Jing Yao, Shanlin Zhou, Zhihua Wei, Peng Zhang, Dongkuan Xu, Maosong Sun, Xing Xie
Big models have achieved revolutionary breakthroughs in the field of AI, but they might also pose potential concerns.
no code implementations • 3 Mar 2024 • Zhu Liu, Cunliang Kong, Ying Liu, Maosong Sun
Large language models have achieved remarkable success in general language understanding tasks.
1 code implementation • 29 Feb 2024 • Yiju Guo, Ganqu Cui, Lifan Yuan, Ning Ding, Jiexin Wang, Huimin Chen, Bowen Sun, Ruobing Xie, Jie zhou, Yankai Lin, Zhiyuan Liu, Maosong Sun
In practice, the multifaceted nature of human preferences inadvertently introduces what is known as the "alignment tax" -a compromise where enhancements in alignment within one objective (e. g., harmlessness) can diminish performance in others (e. g., helpfulness).
no code implementations • 29 Feb 2024 • Shangda Wu, Xu Tan, Zili Wang, Rui Wang, Xiaobing Li, Maosong Sun
Traditional deep learning often overlooks bytes, the basic units of the digital world, where all forms of information and operations are encoded and manipulated in binary format.
1 code implementation • 28 Feb 2024 • Weize Chen, Chenfei Yuan, Jiarui Yuan, Yusheng Su, Chen Qian, Cheng Yang, Ruobing Xie, Zhiyuan Liu, Maosong Sun
Natural language (NL) has long been the predominant format for human cognition and communication, and by extension, has been similarly pivotal in the development and application of Large Language Models (LLMs).
no code implementations • 27 Feb 2024 • Xiaolong Wang, Yile Wang, Yuanchi Zhang, Fuwen Luo, Peng Li, Maosong Sun, Yang Liu
Based on the characteristics of the tasks and the strong dialogue-generation capabilities of LLMs, we propose RiC (Reasoning in Conversation), a method that focuses on solving subjective tasks through dialogue simulation.
no code implementations • 26 Feb 2024 • Jingsi Yu, Cunliang Kong, Liner Yang, Meishan Zhang, Lin Zhu, Yujie Wang, Haozhe Lin, Maosong Sun, Erhong Yang
Sentence Pattern Structure (SPS) parsing is a syntactic analysis method primarily employed in language teaching. Existing SPS parsers rely heavily on textbook corpora for training, lacking cross-domain capability. To overcome this constraint, this paper proposes an innovative approach leveraging large language models (LLMs) within a self-training framework.
1 code implementation • 26 Feb 2024 • Qinyu Luo, Yining Ye, Shihao Liang, Zhong Zhang, Yujia Qin, Yaxi Lu, Yesai Wu, Xin Cong, Yankai Lin, Yingli Zhang, Xiaoyin Che, Zhiyuan Liu, Maosong Sun
Generative models have demonstrated considerable potential in software engineering, particularly in tasks such as code generation and debugging.
no code implementations • 23 Feb 2024 • Yufei Huang, Shengding Hu, Xu Han, Zhiyuan Liu, Maosong Sun
Recent studies have uncovered intriguing phenomena in deep learning, such as grokking, double descent, and emergent abilities in large language models, which challenge human intuition and are crucial for a deeper understanding of neural models.
1 code implementation • 21 Feb 2024 • Meng Xu, Shuo Wang, Liner Yang, Haoyu Wang, Zhenghao Liu, Cunliang Kong, Yun Chen, Yang Liu, Maosong Sun, Erhong Yang
We evaluate several representative multilingual LLMs on the proposed OMGEval, which we believe will provide a valuable reference for the community to further understand and improve the multilingual capability of LLMs.
1 code implementation • 21 Feb 2024 • Xinrong Zhang, Yingfa Chen, Shengding Hu, Zihang Xu, JunHao Chen, Moo Khai Hao, Xu Han, Zhen Leng Thai, Shuo Wang, Zhiyuan Liu, Maosong Sun
Processing and reasoning over long contexts is crucial for many practical applications of Large Language Models (LLMs), such as document comprehension and agent construction.
1 code implementation • 21 Feb 2024 • Weilin Zhao, Yuxiang Huang, Xu Han, Chaojun Xiao, Zhiyuan Liu, Maosong Sun
In this paper, we introduce Ouroboros, which constructs a phrase candidate pool from the verification process of LLMs to provide candidates for draft generation of the small model.
1 code implementation • 21 Feb 2024 • Chenyang Song, Xu Han, Zhengyan Zhang, Shengding Hu, Xiyu Shi, Kuai Li, Chen Chen, Zhiyuan Liu, Guangli Li, Tao Yang, Maosong Sun
Some recent efforts have explored introducing ReLU or its variants as the substitutive activation function to help LLMs achieve activation sparsity and inference acceleration, but few can simultaneously obtain high sparsity and comparable model performance.
no code implementations • 21 Feb 2024 • Fuwen Luo, Chi Chen, Zihao Wan, Zhaolu Kang, Qidong Yan, Yingjie Li, Xiaolong Wang, Siyu Wang, Ziyue Wang, Xiaoyue Mi, Peng Li, Ning Ma, Maosong Sun, Yang Liu
Multimodal large language models (MLLMs) have demonstrated promising results in a variety of tasks that combine vision and language.
1 code implementation • 21 Feb 2024 • Chaoqun He, Renjie Luo, Yuzhuo Bai, Shengding Hu, Zhen Leng Thai, Junhao Shen, Jinyi Hu, Xu Han, Yujie Huang, Yuxiang Zhang, Jie Liu, Lei Qi, Zhiyuan Liu, Maosong Sun
Notably, the best-performing model, GPT-4V, attains an average score of 17. 23% on OlympiadBench, with a mere 11. 28% in physics, highlighting the benchmark rigor and the intricacy of physical reasoning.
1 code implementation • 20 Feb 2024 • Chi Chen, Yiyang Du, Zheng Fang, Ziyue Wang, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu
In this paper, we propose a new paradigm through the model composition of existing MLLMs to create a new model that retains the modal understanding capabilities of each original model.
1 code implementation • 19 Feb 2024 • Ziyue Wang, Chi Chen, Yiqi Zhu, Fuwen Luo, Peng Li, Ming Yan, Ji Zhang, Fei Huang, Maosong Sun, Yang Liu
With the bloom of Large Language Models (LLMs), Multimodal Large Language Models (MLLMs) that incorporate LLMs with pre-trained vision models have recently demonstrated impressive performance across diverse vision-language tasks.
1 code implementation • 19 Feb 2024 • Yuanchi Zhang, Yile Wang, Zijun Liu, Shuo Wang, Xiaolong Wang, Peng Li, Maosong Sun, Yang Liu
While large language models (LLMs) have been pre-trained on multilingual corpora, their performance still lags behind in most languages compared to a few resource-rich languages.
1 code implementation • 18 Feb 2024 • Zhiyu Yang, Zihan Zhou, Shuo Wang, Xin Cong, Xu Han, Yukun Yan, Zhenghao Liu, Zhixing Tan, Pengyuan Liu, Dong Yu, Zhiyuan Liu, Xiaodong Shi, Maosong Sun
Scientific data visualization plays a crucial role in research by enabling the direct display of complex information and assisting researchers in identifying implicit patterns.
no code implementations • 18 Feb 2024 • Hanqing Wang, Bowen Ping, Shuo Wang, Xu Han, Yun Chen, Zhiyuan Liu, Maosong Sun
Most prior works on LoRA combination primarily rely on task-level weights for each involved LoRA, making different examples and tokens share the same LoRA weights.
1 code implementation • 14 Feb 2024 • Cheng Qian, Bingxiang He, Zhong Zhuang, Jia Deng, Yujia Qin, Xin Cong, Zhong Zhang, Jie zhou, Yankai Lin, Zhiyuan Liu, Maosong Sun
Current language model-driven agents often lack mechanisms for effective user participation, which is crucial given the vagueness commonly found in user instructions.
1 code implementation • 12 Feb 2024 • Jiarui Zhang, Jinyi Hu, Mahyar Khayatkhoei, Filip Ilievski, Maosong Sun
Multimodal Large Language Models (MLLMs) have recently shown remarkable perceptual capability in answering visual questions, however, little is known about the limits of their perception.
no code implementations • 7 Feb 2024 • Chaojun Xiao, Pengle Zhang, Xu Han, Guangxuan Xiao, Yankai Lin, Zhengyan Zhang, Zhiyuan Liu, Song Han, Maosong Sun
To alleviate these issues, existing efforts employ sliding attention windows and discard distant tokens to achieve the processing of extremely long sequences.
1 code implementation • 7 Feb 2024 • Haoyu Wang, Shuo Wang, Yukun Yan, Xujia Wang, Zhiyu Yang, Yuzhuang Xu, Zhenghao Liu, Liner Yang, Ning Ding, Xu Han, Zhiyuan Liu, Maosong Sun
Different from previous works that simply translate English instructions, we consider both the language-specific and language-agnostic abilities of LLMs.
no code implementations • 6 Feb 2024 • Zhengyan Zhang, Yixin Song, Guanghui Yu, Xu Han, Yankai Lin, Chaojun Xiao, Chenyang Song, Zhiyuan Liu, Zeyu Mi, Maosong Sun
To find the most efficient activation function for sparse computation, we propose a systematic framework to examine the sparsity of LLMs from three aspects: the trade-off between sparsity and performance, the predictivity of sparsity, and the hardware affinity.
no code implementations • 5 Feb 2024 • Junjie Fang, Likai Tang, Hongzhe Bi, Yujia Qin, Si Sun, Zhenyu Li, Haolun Li, Yongjian Li, Xin Cong, Yukun Yan, Xiaodong Shi, Sen Song, Yankai Lin, Zhiyuan Liu, Maosong Sun
Although there exist various methods devoted to enhancing the long-context processing ability of large language models (LLMs), they are developed in an isolated manner and lack systematic analysis and integration of their strengths, hindering further developments.
no code implementations • 25 Jan 2024 • Cheng Qian, Shihao Liang, Yujia Qin, Yining Ye, Xin Cong, Yankai Lin, Yesai Wu, Zhiyuan Liu, Maosong Sun
This paper introduces Investigate-Consolidate-Exploit (ICE), a novel strategy for enhancing the adaptability and flexibility of AI agents through inter-task self-evolution.
1 code implementation • 9 Jan 2024 • Runchu Tian, Yining Ye, Yujia Qin, Xin Cong, Yankai Lin, Yinxu Pan, Yesai Wu, Zhiyuan Liu, Maosong Sun
Previous evaluations of LLMs' debugging ability are significantly limited by the risk of data leakage, the scale of the dataset, and the variety of tested bugs.
1 code implementation • 28 Dec 2023 • Chen Qian, Yufan Dang, Jiahao Li, Wei Liu, Weize Chen, Cheng Yang, Zhiyuan Liu, Maosong Sun
Recent advancements in large language models (LLMs) have brought significant changes to various domains, especially through LLM-driven autonomous agents.
no code implementations • 28 Dec 2023 • Bohan Lyu, Xin Cong, Heyang Yu, Pan Yang, Yujia Qin, Yining Ye, Yaxi Lu, Zhong Zhang, Yukun Yan, Yankai Lin, Zhiyuan Liu, Maosong Sun
As GitHub has hosted a multitude of repositories which can be seen as a good resource for tools, a promising solution is that LLM-based agents can autonomously integrate the repositories in GitHub according to the user queries to extend their tool set.
2 code implementations • 1 Dec 2023 • Tianyu Yu, Yuan YAO, Haoye Zhang, Taiwen He, Yifeng Han, Ganqu Cui, Jinyi Hu, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun, Tat-Seng Chua
Multimodal Large Language Models (MLLMs) have recently demonstrated impressive capabilities in multimodal understanding, reasoning, and interaction.
1 code implementation • 20 Nov 2023 • Ning Ding, Xingtai Lv, Qiaosen Wang, Yulin Chen, BoWen Zhou, Zhiyuan Liu, Maosong Sun
Recognizing the need for more flexible adaptation, we extend the methodology of LoRA to an innovative approach we call sparse low-rank adaptation (SoRA) that enables dynamic adjustments to the intrinsic rank during the adaptation process.
1 code implementation • 2 Nov 2023 • Yining Ye, Xin Cong, Shizuo Tian, Jiannan Cao, Hao Wang, Yujia Qin, Yaxi Lu, Heyang Yu, Huadong Wang, Yankai Lin, Zhiyuan Liu, Maosong Sun
Empirical experiments are conducted to detail its construction and execution procedure of workflow, showcasing the feasibility of APA, unveiling the possibility of a new paradigm of automation driven by agents.
1 code implementation • 24 Oct 2023 • Qingquan Li, Yiran Hu, Feng Yao, Chaojun Xiao, Zhiyuan Liu, Maosong Sun, Weixing Shen
Furthermore, the case similarities are typically measured solely by the textual semantics of the fact descriptions, which may fail to capture the full complexity of legal cases from the perspective of legal knowledge.
1 code implementation • 24 Oct 2023 • Chaojun Xiao, Yuqi Luo, Wenbin Zhang, Pengle Zhang, Xu Han, Yankai Lin, Zhengyan Zhang, Ruobing Xie, Zhiyuan Liu, Maosong Sun, Jie zhou
Pre-trained language models (PLMs) have achieved remarkable results on NLP tasks but at the expense of huge parameter sizes and the consequent computational costs.
no code implementations • 19 Oct 2023 • Weize Chen, Xiaoyue Xu, Xu Han, Yankai Lin, Ruobing Xie, Zhiyuan Liu, Maosong Sun, Jie zhou
Parameter-shared pre-trained language models (PLMs) have emerged as a successful approach in resource-constrained environments, enabling substantial reductions in model storage and memory costs without significant performance compromise.
no code implementations • 8 Oct 2023 • Yile Wang, Peng Li, Maosong Sun, Yang Liu
Large language models (LLMs) have shown superior performance without task-specific fine-tuning.
no code implementations • 5 Oct 2023 • Shengding Hu, Xin Liu, Xu Han, Xinrong Zhang, Chaoqun He, Weilin Zhao, Yankai Lin, Ning Ding, Zebin Ou, Guoyang Zeng, Zhiyuan Liu, Maosong Sun
With PassUntil, we conduct a quantitative investigation into the scaling law of task performance.
2 code implementations • 2 Oct 2023 • Ganqu Cui, Lifan Yuan, Ning Ding, Guanming Yao, Wei Zhu, Yuan Ni, Guotong Xie, Zhiyuan Liu, Maosong Sun
However, the scarcity of diverse, naturalistic datasets of human preferences on LLM outputs at scale poses a great challenge to RLHF as well as feedback learning research within the open-source community.
2 code implementations • 1 Oct 2023 • Tianyu Yu, Jinyi Hu, Yuan YAO, Haoye Zhang, Yue Zhao, Chongyi Wang, Shan Wang, Yinxv Pan, Jiao Xue, Dahai Li, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun
The capabilities of MLLMs depend on two crucial factors: the model architecture to facilitate the feature alignment of visual modules and large language models; the multimodal instruction tuning datasets for human instruction following.
1 code implementation • 26 Sep 2023 • Chenyang Song, Xu Han, Zheni Zeng, Kuai Li, Chen Chen, Zhiyuan Liu, Maosong Sun, Tao Yang
First, Static ConPET can adapt former continual learning methods originally designed for relatively smaller models to LLMs through PET and a dynamic replay strategy, which largely reduces the tuning costs and alleviates the over-fitting and forgetting issue.
no code implementations • 19 Sep 2023 • Kunlun Zhu, Shihao Liang, Xu Han, Zhi Zheng, Guoyang Zeng, Zhiyuan Liu, Maosong Sun
Recent years have witnessed the success of question answering (QA), especially its potential to be a foundation paradigm for tackling diverse NLP tasks.
1 code implementation • 25 Aug 2023 • Chi Chen, Ruoyu Qin, Fuwen Luo, Xiaoyue Mi, Peng Li, Maosong Sun, Yang Liu
However, existing visual instruction tuning methods only utilize image-language instruction data to align the language and image modalities, lacking a more fine-grained cross-modal alignment.
no code implementations • 24 Aug 2023 • Yining Ye, Xin Cong, Shizuo Tian, Yujia Qin, Chong Liu, Yankai Lin, Zhiyuan Liu, Maosong Sun
Central to the development of rationality is the construction of an internalized utility judgment, capable of assigning numerical utilities to each decision.
2 code implementations • 23 Aug 2023 • Jinyi Hu, Yuan YAO, Chongyi Wang, Shan Wang, Yinxu Pan, Qianyu Chen, Tianyu Yu, Hanghao Wu, Yue Zhao, Haoye Zhang, Xu Han, Yankai Lin, Jiao Xue, Dahai Li, Zhiyuan Liu, Maosong Sun
Building a competitive counterpart in other languages is highly challenging due to the low-resource nature of non-English multimodal data (i. e., lack of large-scale, high-quality image-text data).
1 code implementation • 21 Aug 2023 • Weize Chen, Yusheng Su, Jingwei Zuo, Cheng Yang, Chenfei Yuan, Chi-Min Chan, Heyang Yu, Yaxi Lu, Yi-Hsin Hung, Chen Qian, Yujia Qin, Xin Cong, Ruobing Xie, Zhiyuan Liu, Maosong Sun, Jie zhou
Autonomous agents empowered by Large Language Models (LLMs) have undergone significant improvements, enabling them to generalize across a broad spectrum of tasks.
1 code implementation • 31 Jul 2023 • Yujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu, Lan Yan, Yaxi Lu, Yankai Lin, Xin Cong, Xiangru Tang, Bill Qian, Sihan Zhao, Lauren Hong, Runchu Tian, Ruobing Xie, Jie zhou, Mark Gerstein, Dahai Li, Zhiyuan Liu, Maosong Sun
Based on ToolBench, we fine-tune LLaMA to obtain an LLM ToolLLaMA, and equip it with a neural API retriever to recommend appropriate APIs for each instruction.
Ranked #3 on Trajectory Planning on ToolBench
1 code implementation • 28 Jul 2023 • Shihao Liang, Runchu Tian, Kunlun Zhu, Yujia Qin, Huadong Wang, Xin Cong, Zhiyuan Liu, Xiaojiang Liu, Maosong Sun
Instruction tuning has emerged as a promising approach to enhancing large language models in following human instructions.
1 code implementation • 16 Jul 2023 • Chen Qian, Xin Cong, Wei Liu, Cheng Yang, Weize Chen, Yusheng Su, Yufan Dang, Jiahao Li, Juyuan Xu, Dahai Li, Zhiyuan Liu, Maosong Sun
At the core of this paradigm lies ChatDev, a virtual chat-powered software development company that mirrors the established waterfall model, meticulously dividing the development process into four distinct chronological stages: designing, coding, testing, and documenting.
no code implementations • 15 Jul 2023 • Weilin Zhao, Yuxiang Huang, Xu Han, Zhiyuan Liu, Zhengyan Zhang, Maosong Sun
Parameter-efficient tuning (PET) has been widely explored in recent years because it tunes much fewer parameters (PET modules) than full-parameter fine-tuning (FT) while still stimulating sufficient knowledge from large language models (LLMs) for downstream tasks.
1 code implementation • 5 Jul 2023 • Shengding Hu, Ning Ding, Weilin Zhao, Xingtai Lv, Zhen Zhang, Zhiyuan Liu, Maosong Sun
The scale of large pre-trained models (PTMs) poses significant challenges in adapting to downstream tasks due to the high optimization overhead and storage costs associated with full-parameter fine-tuning.
1 code implementation • 5 Jul 2023 • Shengding Hu, Yifan Luo, Huadong Wang, Xingyi Cheng, Zhiyuan Liu, Maosong Sun
In this paper, we find that the PLMs already possess the knowledge required to rebut such questions, and the key is how to activate the knowledge.
1 code implementation • 21 Jun 2023 • Zheni Zeng, Bangchen Yin, Shipeng Wang, Jiarui Liu, Cheng Yang, Haishen Yao, Xingzhi Sun, Maosong Sun, Guotong Xie, Zhiyuan Liu
Natural language is expected to be a key medium for various human-machine interactions in the era of large language models.
1 code implementation • 7 Jun 2023 • Lifan Yuan, Yangyi Chen, Ganqu Cui, Hongcheng Gao, Fangyuan Zou, Xingyi Cheng, Heng Ji, Zhiyuan Liu, Maosong Sun
Then we introduce BOSS, a Benchmark suite for Out-of-distribution robustneSS evaluation covering 5 tasks and 20 datasets.
1 code implementation • 4 Jun 2023 • Yusheng Su, Chi-Min Chan, Jiali Cheng, Yujia Qin, Yankai Lin, Shengding Hu, Zonghan Yang, Ning Ding, Xingzhi Sun, Guotong Xie, Zhiyuan Liu, Maosong Sun
Our investigations reveal that model scaling (1) mitigates the effects of the positions of tunable parameters on performance, and (2) enables tuning methods to achieve performance comparable to full-parameter fine-tuning by optimizing fewer tunable parameters.
1 code implementation • 29 May 2023 • Yangyi Chen, Hongcheng Gao, Ganqu Cui, Lifan Yuan, Dehan Kong, Hanlu Wu, Ning Shi, Bo Yuan, Longtao Huang, Hui Xue, Zhiyuan Liu, Maosong Sun, Heng Ji
In our experiments, we conduct a robustness evaluation of RoBERTa models to demonstrate the effectiveness of our evaluation framework, and further show the rationality of each component in the framework.
1 code implementation • 28 May 2023 • Zhengyan Zhang, Zhiyuan Zeng, Yankai Lin, Chaojun Xiao, Xiaozhi Wang, Xu Han, Zhiyuan Liu, Ruobing Xie, Maosong Sun, Jie zhou
In analogy to human brains, we consider two main characteristics of modularity: (1) functional specialization of neurons: we evaluate whether each neuron is mainly specialized in a certain function, and find that the answer is yes.
1 code implementation • 28 May 2023 • Chaojun Xiao, Zhengyan Zhang, Xu Han, Chi-Min Chan, Yankai Lin, Zhiyuan Liu, Xiangyang Li, Zhonghua Li, Zhao Cao, Maosong Sun
By inserting document plugins into the backbone PTM for downstream tasks, we can encode a document one time to handle multiple tasks, which is more efficient than conventional encoding-task coupling methods that simultaneously encode documents and input queries using task-specific encoders.
1 code implementation • 28 May 2023 • Zhengyan Zhang, Zhiyuan Zeng, Yankai Lin, Huadong Wang, Deming Ye, Chaojun Xiao, Xu Han, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Experimental results on three knowledge-driven NLP tasks show that existing injection methods are not suitable for the new paradigm, while map-tuning effectively improves the performance of downstream models.
1 code implementation • 28 May 2023 • Weize Chen, Xu Han, Yankai Lin, Zhiyuan Liu, Maosong Sun, Jie zhou
Since it is non-trivial to directly model the intermediate states and design a running cost function, we propose to use latent stochastic bridges to regularize the intermediate states and use the regularization as the running cost of PETs.
no code implementations • 24 May 2023 • Chi Chen, Peng Li, Maosong Sun, Yang Liu
Weakly supervised vision-and-language pre-training (WVLP), which learns cross-modal representations with limited cross-modal supervision, has been shown to effectively reduce the data cost of pre-training while maintaining decent performance on downstream tasks.
1 code implementation • 23 May 2023 • Ning Ding, Yulin Chen, Bokai Xu, Yujia Qin, Zhi Zheng, Shengding Hu, Zhiyuan Liu, Maosong Sun, BoWen Zhou
Fine-tuning on instruction data has been widely validated as an effective practice for implementing chat language models like ChatGPT.
no code implementations • 19 May 2023 • Jinyi Hu, Xu Han, Xiaoyuan Yi, Yutong Chen, Wenhao Li, Zhiyuan Liu, Maosong Sun
IAP optimizes only a separate Chinese text encoder with all other parameters fixed to align Chinese semantics space to the English one in CLIP.
1 code implementation • NeurIPS 2023 • Yuzhen Huang, Yuzhuo Bai, Zhihao Zhu, Junlei Zhang, Jinghan Zhang, Tangjun Su, Junteng Liu, Chuancheng Lv, Yikai Zhang, Jiayi Lei, Yao Fu, Maosong Sun, Junxian He
We present C-Eval, the first comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context.
1 code implementation • 15 May 2023 • Yujia Qin, Cheng Qian, Xu Han, Yankai Lin, Huadong Wang, Ruobing Xie, Zhiyuan Liu, Maosong Sun, Jie zhou
In pilot studies, we find that after continual pre-training, the upgraded PLM remains compatible with the outdated adapted weights to some extent.
1 code implementation • 11 May 2023 • Yujia Qin, Zihan Cai, Dian Jin, Lan Yan, Shihao Liang, Kunlun Zhu, Yankai Lin, Xu Han, Ning Ding, Huadong Wang, Ruobing Xie, Fanchao Qi, Zhiyuan Liu, Maosong Sun, Jie zhou
We recruit annotators to search for relevant information using our interface and then answer questions.
no code implementations • 2 May 2023 • Deming Ye, Yankai Lin, Zhengyan Zhang, Maosong Sun
In this paper, we propose a UNified knowledge inTERface, UNTER, to provide a unified perspective to exploit both structured knowledge and unstructured knowledge.
2 code implementations • 21 Apr 2023 • Shangda Wu, Dingyao Yu, Xu Tan, Maosong Sun
We introduce CLaMP: Contrastive Language-Music Pre-training, which learns cross-modal representations between natural language and symbolic music using a music encoder and a text encoder trained jointly with a contrastive loss.
3 code implementations • 17 Apr 2023 • Yujia Qin, Shengding Hu, Yankai Lin, Weize Chen, Ning Ding, Ganqu Cui, Zheni Zeng, Yufei Huang, Chaojun Xiao, Chi Han, Yi Ren Fung, Yusheng Su, Huadong Wang, Cheng Qian, Runchu Tian, Kunlun Zhu, Shihao Liang, Xingyu Shen, Bokai Xu, Zhen Zhang, Yining Ye, Bowen Li, Ziwei Tang, Jing Yi, Yuzhang Zhu, Zhenning Dai, Lan Yan, Xin Cong, Yaxi Lu, Weilin Zhao, Yuxiang Huang, Junxi Yan, Xu Han, Xian Sun, Dahai Li, Jason Phang, Cheng Yang, Tongshuang Wu, Heng Ji, Zhiyuan Liu, Maosong Sun
Considering the lack of a systematic tool learning evaluation in prior works, we experiment with 18 representative tools and show the potential of current foundation models in skillfully utilizing tools.
1 code implementation • 14 Feb 2023 • Chenglei Si, Zhengyan Zhang, Yingfa Chen, Xiaozhi Wang, Zhiyuan Liu, Maosong Sun
In order to fill this important gap, we construct READIN: a Chinese multi-task benchmark with REalistic And Diverse Input Noises.
1 code implementation • 19 Dec 2022 • Xuancheng Huang, Zijun Liu, Peng Li, Tao Li, Maosong Sun, Yang Liu
Recently, multi-aspect controllable text generation that controls the generated text in multiple aspects (e. g., sentiment, topic, and keywords) has attracted increasing attention.
1 code implementation • 18 Dec 2022 • Yuanchi Zhang, Peng Li, Maosong Sun, Yang Liu
While many parallel corpora are not publicly accessible for data copyright, data privacy and competitive differentiation reasons, trained translation models are increasingly available on open platforms.
2 code implementations • 16 Dec 2022 • Ganqu Cui, Wentao Li, Ning Ding, Longtao Huang, Zhiyuan Liu, Maosong Sun
With the evergrowing sizes of pre-trained models (PTMs), it has been an emerging practice to only provide the inference APIs for users, namely model-as-a-service (MaaS) setting.
1 code implementation • 22 Nov 2022 • Yuan YAO, Tianyu Yu, Ao Zhang, Mengdi Li, Ruobing Xie, Cornelius Weber, Zhiyuan Liu, Hai-Tao Zheng, Stefan Wermter, Tat-Seng Chua, Maosong Sun
In this work, we present CLEVER, which formulates CKE as a distantly supervised multi-instance learning problem, where models learn to summarize commonsense relations from a bag of images about an entity pair without any human annotation on image instances.
1 code implementation • 21 Nov 2022 • Shangda Wu, Maosong Sun
Benefiting from large-scale datasets and pre-trained models, the field of generative models has recently gained significant momentum.
1 code implementation • 14 Nov 2022 • Wenhao Li, Xiaoyuan Yi, Jinyi Hu, Maosong Sun, Xing Xie
In this work, we dig into the intrinsic mechanism of this problem and found that sparser attention values in Transformer could improve diversity.
1 code implementation • 13 Nov 2022 • Yufei Huang, Yujia Qin, Huadong Wang, Yichun Yin, Maosong Sun, Zhiyuan Liu, Qun Liu
Inspired by these observations, we propose Fast Prompt Tuning (FPT), which starts by conducting PT using a small-scale partial PLM, and then progressively expands its depth and width until the full-model size.
1 code implementation • NIPS 2022 • Shengding Hu, Zhen Zhang, Ning Ding, Yadao Wang, Yasheng Wang, Zhiyuan Liu, Maosong Sun
Generally, DT methods exquisitely design delta modules (DT modules) which could be applied to arbitrary fine-grained positions inside PTMs.
1 code implementation • 25 Oct 2022 • Yujia Qin, Cheng Qian, Jing Yi, Weize Chen, Yankai Lin, Xu Han, Zhiyuan Liu, Maosong Sun, Jie zhou
(3) How does the PLM's task knowledge change along the path connecting two minima?
1 code implementation • 24 Oct 2022 • Jing Yi, Weize Chen, Yujia Qin, Yankai Lin, Ning Ding, Xu Han, Zhiyuan Liu, Maosong Sun, Jie zhou
To fathom the mystery, we hypothesize that the adaptations of different DETs could all be reparameterized as low-dimensional optimizations in a unified optimization subspace, which could be found by jointly decomposing independent solutions of different DETs.
no code implementations • 22 Oct 2022 • Jinyi Hu, Xiaoyuan Yi, Wenhao Li, Maosong Sun, Xing Xie
We demonstrate that TRACE could enhance the entanglement of each segment and preceding latent variables and deduce a non-zero lower bound of the KL term, providing a theoretical guarantee of generation diversity.
1 code implementation • 19 Oct 2022 • Yangyi Chen, Hongcheng Gao, Ganqu Cui, Fanchao Qi, Longtao Huang, Zhiyuan Liu, Maosong Sun
We discuss the deficiencies in previous work and propose our suggestions that the research on the Security-oriented adversarial NLP (SoadNLP) should: (1) evaluate their methods on security tasks to demonstrate the real-world concerns; (2) consider real-world attackers' goals, instead of developing impractical methods.
1 code implementation • COLING 2022 • Zichun Yu, Tianyu Gao, Zhengyan Zhang, Yankai Lin, Zhiyuan Liu, Maosong Sun, Jie zhou
Prompting, which casts downstream applications as language modeling tasks, has shown to be sample efficient compared to standard fine-tuning with pre-trained models.
1 code implementation • NAACL 2022 • Jinyi Hu, Xiaoyuan Yi, Wenhao Li, Maosong Sun, Xing Xie
The past several years have witnessed Variational Auto-Encoder's superiority in various text generation tasks.
no code implementations • 19 Jun 2022 • Zhen Li, Xiting Wang, Weikai Yang, Jing Wu, Zhengyan Zhang, Zhiyuan Liu, Maosong Sun, HUI ZHANG, Shixia Liu
The rapid development of deep natural language processing (NLP) models for text classification has led to an urgent need for a unified understanding of these models proposed individually.
1 code implementation • 17 Jun 2022 • Ganqu Cui, Lifan Yuan, Bingxiang He, Yangyi Chen, Zhiyuan Liu, Maosong Sun
However, we highlight two issues in previous backdoor learning evaluations: (1) The differences between real-world scenarios (e. g. releasing poisoned datasets or models) are neglected, and we argue that each scenario has its own constraints and concerns, thus requires specific evaluation protocols; (2) The evaluation metrics only consider whether the attacks could flip the models' predictions on poisoned samples and retain performances on benign samples, but ignore that poisoned samples should also be stealthy and semantic-preserving.
no code implementations • 15 Jun 2022 • Shengding Hu, Zhen Zhang, Ning Ding, Yadao Wang, Yasheng Wang, Zhiyuan Liu, Maosong Sun
The searched structures preserve more than 99\% fine-tuning performance with 0. 01\% trainable parameters.
1 code implementation • Findings (ACL) 2022 • Yuan YAO, Bowen Dong, Ao Zhang, Zhengyan Zhang, Ruobing Xie, Zhiyuan Liu, Leyu Lin, Maosong Sun, Jianyong Wang
Recent works have shown promising results of prompt tuning in stimulating pre-trained language models (PLMs) for natural language processing (NLP) tasks.
1 code implementation • 23 May 2022 • Yuan YAO, Qianyu Chen, Ao Zhang, Wei Ji, Zhiyuan Liu, Tat-Seng Chua, Maosong Sun
We show that PEVL enables state-of-the-art performance of detector-free VLP models on position-sensitive tasks such as referring expression comprehension and phrase grounding, and also improves the performance on position-insensitive tasks with grounded inputs.
Ranked #1 on Visual Commonsense Reasoning on VCR (Q-AR) test
1 code implementation • 23 May 2022 • Shuo Wang, Peng Li, Zhixing Tan, Zhaopeng Tu, Maosong Sun, Yang Liu
In this work, we propose a template-based method that can yield results with high translation quality and match accuracy and the inference speed of our method is comparable with unconstrained NMT models.
no code implementations • 12 May 2022 • Shangda Wu, Maosong Sun
In recent years, there has been a growing interest in the development of language models capable of generating text with controllable attributes.
1 code implementation • 10 May 2022 • Jiafeng Liu, Yuanliang Dong, Zehua Cheng, Xinran Zhang, Xiaobing Li, Feng Yu, Maosong Sun
In this work, we propose a permutation invariant language model, SymphonyNet, as a solution for symbolic symphony music generation.
Ranked #1 on Audio Generation on Symphony music
no code implementations • 26 Mar 2022 • Sha Yuan, Hanyu Zhao, Shuai Zhao, Jiahong Leng, Yangxiao Liang, Xiaozhi Wang, Jifan Yu, Xin Lv, Zhou Shao, Jiaao He, Yankai Lin, Xu Han, Zhenghao Liu, Ning Ding, Yongming Rao, Yizhao Gao, Liang Zhang, Ming Ding, Cong Fang, Yisen Wang, Mingsheng Long, Jing Zhang, Yinpeng Dong, Tianyu Pang, Peng Cui, Lingxiao Huang, Zheng Liang, HuaWei Shen, HUI ZHANG, Quanshi Zhang, Qingxiu Dong, Zhixing Tan, Mingxuan Wang, Shuo Wang, Long Zhou, Haoran Li, Junwei Bao, Yingwei Pan, Weinan Zhang, Zhou Yu, Rui Yan, Chence Shi, Minghao Xu, Zuobai Zhang, Guoqiang Wang, Xiang Pan, Mengjie Li, Xiaoyu Chu, Zijun Yao, Fangwei Zhu, Shulin Cao, Weicheng Xue, Zixuan Ma, Zhengyan Zhang, Shengding Hu, Yujia Qin, Chaojun Xiao, Zheni Zeng, Ganqu Cui, Weize Chen, Weilin Zhao, Yuan YAO, Peng Li, Wenzhao Zheng, Wenliang Zhao, Ziyi Wang, Borui Zhang, Nanyi Fei, Anwen Hu, Zenan Ling, Haoyang Li, Boxi Cao, Xianpei Han, Weidong Zhan, Baobao Chang, Hao Sun, Jiawen Deng, Chujie Zheng, Juanzi Li, Lei Hou, Xigang Cao, Jidong Zhai, Zhiyuan Liu, Maosong Sun, Jiwen Lu, Zhiwu Lu, Qin Jin, Ruihua Song, Ji-Rong Wen, Zhouchen Lin, LiWei Wang, Hang Su, Jun Zhu, Zhifang Sui, Jiajun Zhang, Yang Liu, Xiaodong He, Minlie Huang, Jian Tang, Jie Tang
With the rapid development of deep learning, training Big Models (BMs) for multiple downstream tasks becomes a popular paradigm.
2 code implementations • 22 Mar 2022 • Ao Zhang, Yuan YAO, Qianyu Chen, Wei Ji, Zhiyuan Liu, Maosong Sun, Tat-Seng Chua
Scene graph generation (SGG) is designed to extract (subject, predicate, object) triplets in images.
Ranked #1 on Predicate Classification on Visual Genome
1 code implementation • Findings (ACL) 2022 • Feng Yao, Chaojun Xiao, Xiaozhi Wang, Zhiyuan Liu, Lei Hou, Cunchao Tu, Juanzi Li, Yun Liu, Weixing Shen, Maosong Sun
However, existing Legal Event Detection (LED) datasets only concern incomprehensive event types and have limited annotated data, which restricts the development of LED methods and their downstream applications.
1 code implementation • 14 Mar 2022 • Ning Ding, Yujia Qin, Guang Yang, Fuchao Wei, Zonghan Yang, Yusheng Su, Shengding Hu, Yulin Chen, Chi-Min Chan, Weize Chen, Jing Yi, Weilin Zhao, Xiaozhi Wang, Zhiyuan Liu, Hai-Tao Zheng, Jianfei Chen, Yang Liu, Jie Tang, Juanzi Li, Maosong Sun
This necessitates a new branch of research focusing on the parameter-efficient adaptation of PLMs, dubbed as delta tuning in this paper.
1 code implementation • Findings (ACL) 2022 • Fanchao Qi, Chuancheng Lv, Zhiyuan Liu, Xiaojun Meng, Maosong Sun, Hai-Tao Zheng
In this paper, we utilize the multilingual synonyms, multilingual glosses and images in BabelNet for SPBS.
1 code implementation • Findings (ACL) 2022 • Yujia Qin, Jiajie Zhang, Yankai Lin, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
We experiment ELLE with streaming data from 5 domains on BERT and GPT.
1 code implementation • ACL 2022 • Deming Ye, Yankai Lin, Peng Li, Maosong Sun, Zhiyuan Liu
Pre-trained language models (PLMs) cannot well recall rich factual knowledge of entities exhibited in large-scale corpora, especially those rare entities.
1 code implementation • ACL 2022 • Fanchao Qi, Yanhui Yang, Jing Yi, Zhili Cheng, Zhiyuan Liu, Maosong Sun
To facilitate the research on this task, we build a large and fully open quote recommendation dataset called QuoteR, which comprises three parts including English, standard Chinese and classical Chinese.
1 code implementation • 17 Feb 2022 • Shangda Wu, Xiaobing Li, Maosong Sun
Melody harmonization has long been closely associated with chorales composed by Johann Sebastian Bach.
1 code implementation • 30 Dec 2021 • Yingying Wang, Cunliang Kong, Liner Yang, Yijun Wang, Xiaorong Lu, Renfen Hu, Shan He, Zhenghao Liu, Yun Chen, Erhong Yang, Maosong Sun
This resource is of great relevance for second language acquisition research, foreign-language teaching, and automatic grammatical error correction.
no code implementations • 27 Dec 2021 • Yuan YAO, Qingxiu Dong, Jian Guan, Boxi Cao, Zhengyan Zhang, Chaojun Xiao, Xiaozhi Wang, Fanchao Qi, Junwei Bao, Jinran Nie, Zheni Zeng, Yuxian Gu, Kun Zhou, Xuancheng Huang, Wenhao Li, Shuhuai Ren, Jinliang Lu, Chengqiang Xu, Huadong Wang, Guoyang Zeng, Zile Zhou, Jiajun Zhang, Juanzi Li, Minlie Huang, Rui Yan, Xiaodong He, Xiaojun Wan, Xin Zhao, Xu sun, Yang Liu, Zhiyuan Liu, Xianpei Han, Erhong Yang, Zhifang Sui, Maosong Sun
We argue that for general-purpose language intelligence evaluation, the benchmark itself needs to be comprehensive and systematic.
1 code implementation • NAACL 2022 • Yusheng Su, Xiaozhi Wang, Yujia Qin, Chi-Min Chan, Yankai Lin, Huadong Wang, Kaiyue Wen, Zhiyuan Liu, Peng Li, Juanzi Li, Lei Hou, Maosong Sun, Jie zhou
To explore whether we can improve PT via prompt transfer, we empirically investigate the transferability of soft prompts across different downstream tasks and PLMs in this work.
2 code implementations • ACL 2022 • Ning Ding, Shengding Hu, Weilin Zhao, Yulin Chen, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun
Prompt-learning has become a new paradigm in modern natural language processing, which directly adapts pre-trained language models (PLMs) to $cloze$-style prediction, autoregressive modeling, or sequence to sequence generation, resulting in promising performances on various tasks.
1 code implementation • 15 Oct 2021 • Yujia Qin, Xiaozhi Wang, Yusheng Su, Yankai Lin, Ning Ding, Jing Yi, Weize Chen, Zhiyuan Liu, Juanzi Li, Lei Hou, Peng Li, Maosong Sun, Jie zhou
In the experiments, we study diverse few-shot NLP tasks and surprisingly find that in a 250-dimensional subspace found with 100 tasks, by only tuning 250 free parameters, we can recover 97% and 83% of the full prompt tuning performance for 100 seen tasks (using different training data) and 20 unseen tasks, respectively, showing great generalization ability of the found intrinsic task subspace.
1 code implementation • 15 Oct 2021 • Yangyi Chen, Fanchao Qi, Hongcheng Gao, Zhiyuan Liu, Maosong Sun
In this paper, we find two simple tricks that can make existing textual backdoor attacks much more harmful.
1 code implementation • EMNLP 2021 • Fanchao Qi, Yangyi Chen, Xurui Zhang, Mukai Li, Zhiyuan Liu, Maosong Sun
In this paper, we make the first attempt to conduct adversarial and backdoor attacks based on text style transfer, which is aimed at altering the style of a sentence while preserving its meaning.
1 code implementation • Findings (ACL) 2022 • Zhengyan Zhang, Yankai Lin, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
In this work, we study the computational patterns of FFNs and observe that most inputs only activate a tiny ratio of neurons of FFNs.
1 code implementation • 24 Sep 2021 • Yuan YAO, Ao Zhang, Zhengyan Zhang, Zhiyuan Liu, Tat-Seng Chua, Maosong Sun
Pre-Trained Vision-Language Models (VL-PTMs) have shown promising capabilities in grounding natural language in image data, facilitating a broad variety of cross-modal tasks.
2 code implementations • ACL 2022 • Deming Ye, Yankai Lin, Peng Li, Maosong Sun
In particular, we propose a neighborhood-oriented packing strategy, which considers the neighbor spans integrally to better model the entity boundary information.
Ranked #1 on Named Entity Recognition (NER) on Few-NERD (SUP)
no code implementations • 27 Aug 2021 • Xinran Zhang, Maosong Sun, Jiafeng Liu, Xiaobing Li
We propose nucleus sampling with randomized head (NS-RH) algorithm, which randomizes the high frequency part ("head") of the predicted distribution, in order to emphasize on the "comparatively low frequency" words.
2 code implementations • ACL 2022 • Shengding Hu, Ning Ding, Huadong Wang, Zhiyuan Liu, Jingang Wang, Juanzi Li, Wei Wu, Maosong Sun
Tuning pre-trained language models (PLMs) with task-specific prompts has been a promising approach for text classification.
no code implementations • 25 Jun 2021 • Shuo Wang, Zhaopeng Tu, Zhixing Tan, Wenxuan Wang, Maosong Sun, Yang Liu
Inspired by the recent progress of large-scale pre-trained language models on machine translation in a limited scenario, we firstly demonstrate that a single language model (LM4MT) can achieve comparable performance with strong encoder-decoder NMT models on standard machine translation benchmarks, using the same training data and similar amount of model parameters.
2 code implementations • 20 Jun 2021 • Zhengyan Zhang, Yuxian Gu, Xu Han, Shengqi Chen, Chaojun Xiao, Zhenbo Sun, Yuan YAO, Fanchao Qi, Jian Guan, Pei Ke, Yanzheng Cai, Guoyang Zeng, Zhixing Tan, Zhiyuan Liu, Minlie Huang, Wentao Han, Yang Liu, Xiaoyan Zhu, Maosong Sun
We present a suite of cost-effective techniques for the use of PLMs to deal with the efficiency issues of pre-training, fine-tuning, and inference.
no code implementations • Findings (ACL) 2021 • Rui Jiao, Zonghan Yang, Maosong Sun, Yang Liu
In this work, we propose alternated training with synthetic and authentic data for NMT.
1 code implementation • ACL 2021 • Fanchao Qi, Yuan YAO, Sophia Xu, Zhiyuan Liu, Maosong Sun
Recent studies show that neural natural language processing (NLP) models are vulnerable to backdoor attacks.
no code implementations • Findings (ACL) 2021 • Shuo Wang, Zhaopeng Tu, Zhixing Tan, Shuming Shi, Maosong Sun, Yang Liu
Language coverage bias, which indicates the content-dependent differences between sentence pairs originating from the source and target languages, is important for neural machine translation (NMT) because the target-original training data is not well exploited in current practice.
1 code implementation • 3 Jun 2021 • Wenhao Li, Fanchao Qi, Maosong Sun, Xiaoyuan Yi, Jiarui Zhang
We hope this dataset can further enhance the study on incorporating deep semantics into the understanding and generation system of Chinese classical poetry.
2 code implementations • 1 Jun 2021 • Chenglei Si, Zhengyan Zhang, Yingfa Chen, Fanchao Qi, Xiaozhi Wang, Zhiyuan Liu, Yasheng Wang, Qun Liu, Maosong Sun
2) Pronunciation-based SubChar tokenizers can encode Chinese homophones into the same transliteration sequences and produce the same tokenization output, hence being robust to homophone typos.
1 code implementation • NAACL 2021 • Kai Zhang, Yuan YAO, Ruobing Xie, Xu Han, Zhiyuan Liu, Fen Lin, Leyu Lin, Maosong Sun
To establish the bidirectional connections between OpenRE and relation hierarchy, we propose the task of open hierarchical relation extraction and present a novel OHRE framework for the task.
1 code implementation • ACL 2022 • Weize Chen, Xu Han, Yankai Lin, Hexu Zhao, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Hyperbolic neural networks have shown great potential for modeling complex data.
1 code implementation • ACL 2021 • Xuancheng Huang, Jingfang Xu, Maosong Sun, Yang Liu
Although directly finetuning pretrained models on MSG tasks and concatenating multiple sources into a single long sequence is regarded as a simple method to transfer pretrained models to MSG tasks, we conjecture that the direct finetuning method leads to catastrophic forgetting and solely relying on pretrained self-attention layers to capture cross-source information is not sufficient.
2 code implementations • NAACL 2022 • Yujia Qin, Yankai Lin, Jing Yi, Jiajie Zhang, Xu Han, Zhengyan Zhang, Yusheng Su, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Specifically, we introduce a pre-training framework named "knowledge inheritance" (KI) and explore how could knowledge distillation serve as auxiliary supervision during pre-training to efficiently learn larger PLMs.
2 code implementations • ACL 2021 • Fanchao Qi, Mukai Li, Yangyi Chen, Zhengyan Zhang, Zhiyuan Liu, Yasheng Wang, Maosong Sun
As far as we know, almost all existing textual backdoor attack methods insert additional contents into normal samples as triggers, which causes the trigger-embedded samples to be detected and the backdoor attacks to be blocked without much effort.
1 code implementation • Findings (ACL) 2021 • Fanchao Qi, Yangyi Chen, Fengyu Wang, Zhiyuan Liu, Xiao Chen, Maosong Sun
We use this method to build an English SKB and a French SKB, and conduct comprehensive evaluations from both intrinsic and extrinsic perspectives.
1 code implementation • NAACL 2021 • Deming Ye, Yankai Lin, Yufei Huang, Maosong Sun
To address this issue, we propose a dynamic token reduction approach to accelerate PLMs' inference, named TR-BERT, which could flexibly adapt the layer number of each token in inference to avoid redundant calculation.
1 code implementation • 24 May 2021 • Xu Han, Weilin Zhao, Ning Ding, Zhiyuan Liu, Maosong Sun
This indicates that PTR is a promising approach to take advantage of both human prior knowledge and PLMs for those complicated classification tasks.
1 code implementation • Findings (ACL) 2021 • Tianyu Gao, Xu Han, Keyue Qiu, Yuzhuo Bai, Zhiyu Xie, Yankai Lin, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Distantly supervised (DS) relation extraction (RE) has attracted much attention in the past few years as it can utilize large-scale auto-labeled data.
1 code implementation • 14 May 2021 • Zhixing Tan, Zeyuan Yang, Meng Zhang, Qun Liu, Maosong Sun, Yang Liu
With the rapid development of artificial intelligence (AI), there is a trend in moving AI applications, such as neural machine translation (NMT), from cloud to mobile devices.
1 code implementation • NAACL 2021 • Zhenghao Liu, Xiaoyuan Yi, Maosong Sun, Liner Yang, Tat-Seng Chua
Grammatical Error Correction (GEC) aims to correct writing errors and help language learners improve their writing skills.
Ranked #1 on Grammatical Error Detection on FCE
1 code implementation • 9 May 2021 • Chaojun Xiao, Xueyu Hu, Zhiyuan Liu, Cunchao Tu, Maosong Sun
Legal artificial intelligence (LegalAI) aims to benefit legal systems with the technology of artificial intelligence, especially natural language processing (NLP).
1 code implementation • ICCV 2021 • Yuan YAO, Ao Zhang, Xu Han, Mengdi Li, Cornelius Weber, Zhiyuan Liu, Stefan Wermter, Maosong Sun
In this work, we propose visual distant supervision, a novel paradigm of visual relation learning, which can train scene graph models without any human-labeled data.
no code implementations • 25 Mar 2021 • Yuzhong Wang, Chaojun Xiao, Shirong Ma, Haoxi Zhong, Cunchao Tu, Tianyang Zhang, Zhiyuan Liu, Maosong Sun
We propose to simulate judges from different groups with legal judgment prediction (LJP) models and measure the judicial inconsistency with the disagreement of the judgment results given by LJP models trained on different groups.
no code implementations • 13 Mar 2021 • Xinran Zhang, Maosong Sun, Jiafeng Liu, Xiaobing Li
In natural language processing (NLP), the semantic similarity task requires large-scale, high-quality human-annotated labels for fine-tuning or evaluation.
no code implementations • 13 Mar 2021 • Xinran Zhang, Maosong Sun, Jiafeng Liu, Xiaobing Li
Traditional stochastic sampling methods only focus on truncating the unreliable "tail" of the distribution, and do not address the "head" part, which we show might contain tedious or even repetitive candidates with high probability that lead to repetition loops.
no code implementations • 22 Feb 2021 • Chaojun Xiao, Ruobing Xie, Yuan YAO, Zhiyuan Liu, Maosong Sun, Xu Zhang, Leyu Lin
Existing sequential recommendation methods rely on large amounts of training data and usually suffer from the data sparsity problem.
1 code implementation • 7 Feb 2021 • Yusheng Su, Xu Han, Yankai Lin, Zhengyan Zhang, Zhiyuan Liu, Peng Li, Jie zhou, Maosong Sun
We then perform contrastive semi-supervised learning on both the retrieved unlabeled and original labeled instances to help PLMs capture crucial task-related semantic features.
no code implementations • 7 Feb 2021 • Zhiyuan Liu, Yankai Lin, Maosong Sun
This book aims to review and present the recent advances of distributed representation learning for NLP, including why representation learning can improve NLP, how representation learning takes part in various important topics of NLP, and what challenges are still not well addressed by distributed representation.
1 code implementation • 30 Jan 2021 • Zhenghao Liu, Kaitao Zhang, Chenyan Xiong, Zhiyuan Liu, Maosong Sun
OpenMatch is a Python-based library that serves for Neural Information Retrieval (Neu-IR) research.
1 code implementation • ICML Workshop AML 2021 • Zhengyan Zhang, Guangxuan Xiao, Yongwei Li, Tian Lv, Fanchao Qi, Zhiyuan Liu, Yasheng Wang, Xin Jiang, Maosong Sun
In this work, we demonstrate the universal vulnerability of PTMs, where fine-tuned PTMs can be easily controlled by backdoor attacks in arbitrary downstream tasks.
1 code implementation • 31 Dec 2020 • Chenglei Si, Zhengyan Zhang, Fanchao Qi, Zhiyuan Liu, Yasheng Wang, Qun Liu, Maosong Sun
In this work, we propose a simple and effective method to cover a much larger proportion of the attack search space, called Adversarial and Mixup Data Augmentation (AMDA).
no code implementations • 31 Dec 2020 • Zhixing Tan, Shuo Wang, Zonghan Yang, Gang Chen, Xuancheng Huang, Maosong Sun, Yang Liu
Machine translation (MT) is an important sub-field of natural language processing that aims to translate natural languages using computers.
1 code implementation • ACL 2021 • Yujia Qin, Yankai Lin, Ryuichi Takanobu, Zhiyuan Liu, Peng Li, Heng Ji, Minlie Huang, Maosong Sun, Jie zhou
Pre-trained Language Models (PLMs) have shown superior performance on various downstream Natural Language Processing (NLP) tasks.
no code implementations • 25 Dec 2020 • Gang Chen, Maosong Sun, Yang Liu
In this work, we propose a method for building a continuous knowledge base (CKB) that can store knowledge imported from multiple, diverse neural networks.
1 code implementation • ACL 2021 • Chi Chen, Maosong Sun, Yang Liu
Word alignment, which aims to align translationally equivalent words between source and target sentences, plays an important role in many natural language processing tasks.
1 code implementation • COLING 2020 • Bowen Dong, Yuan YAO, Ruobing Xie, Tianyu Gao, Xu Han, Zhiyuan Liu, Fen Lin, Leyu Lin, Maosong Sun
Few-shot classification requires classifiers to adapt to new classes with only a few training instances.
1 code implementation • COLING 2020 • Bairu Hou, Fanchao Qi, Yuan Zang, Xurui Zhang, Zhiyuan Liu, Maosong Sun
In this paper, we propose a new unsupervised method for HowNet-based Chinese WSD, which exploits the masked language model task of pre-trained language models.
6 code implementations • 1 Dec 2020 • Zhengyan Zhang, Xu Han, Hao Zhou, Pei Ke, Yuxian Gu, Deming Ye, Yujia Qin, Yusheng Su, Haozhe Ji, Jian Guan, Fanchao Qi, Xiaozhi Wang, Yanan Zheng, Guoyang Zeng, Huanqi Cao, Shengqi Chen, Daixuan Li, Zhenbo Sun, Zhiyuan Liu, Minlie Huang, Wentao Han, Jie Tang, Juanzi Li, Xiaoyan Zhu, Maosong Sun
However, applying GPT-3 to address Chinese NLP tasks is still challenging, as the training corpus of GPT-3 is primarily English, and the parameters are not publicly available.
2 code implementations • EMNLP 2021 • Fanchao Qi, Yangyi Chen, Mukai Li, Yuan YAO, Zhiyuan Liu, Maosong Sun
Nevertheless, there are few studies on defending against textual backdoor attacks.
1 code implementation • EMNLP 2020 • Chaojun Xiao, Yuan YAO, Ruobing Xie, Xu Han, Zhiyuan Liu, Maosong Sun, Fen Lin, Leyu Lin
Distant supervision (DS) has been widely used to generate auto-labeled data for sentence-level relation extraction (RE), which improves RE performance.
no code implementations • 7 Nov 2020 • Zhengyan Zhang, Fanchao Qi, Zhiyuan Liu, Qun Liu, Maosong Sun
To measure the informativeness of attention heads, we train our Single-Shot Meta-Pruner (SMP) with a meta-learning paradigm aiming to maintain the distribution of text representations after pruning.
1 code implementation • Findings of the Association for Computational Linguistics 2020 • Zhenghao Liu, Chenyan Xiong, Zhuyun Dai, Si Sun, Maosong Sun, Zhiyuan Liu
With the epidemic of COVID-19, verifying the scientifically false online information, such as fake news and maliciously fabricated statements, has become crucial.
1 code implementation • NeurIPS 2020 • Wangchunshu Zhou, Jinyi Hu, HANLIN ZHANG, Xiaodan Liang, Maosong Sun, Chenyan Xiong, Jian Tang
In this paper, we develop a general framework for interpretable natural language understanding that requires only a small set of human annotated explanations for training.
1 code implementation • EMNLP 2020 • Hao Peng, Tianyu Gao, Xu Han, Yankai Lin, Peng Li, Zhiyuan Liu, Maosong Sun, Jie zhou
We find that (i) while context is the main source to support the predictions, RE models also heavily rely on the information from entity mentions, most of which is type information, and (ii) existing datasets may leak shallow heuristics via entity mentions and thus contribute to the high performance on RE benchmarks.
Ranked #24 on Relation Extraction on TACRED
1 code implementation • EMNLP 2020 • Fanchao Qi, Lei Zhang, Yanhui Yang, Zhiyuan Liu, Maosong Sun
A reverse dictionary takes descriptions of words as input and outputs words semantically matching the input descriptions.
no code implementations • EMNLP 2020 • Xu Han, Yuzhuo Bai, Keyue Qiu, Zhiyuan Liu, Maosong Sun
Oracle bone script (OBS) is the earliest known ancient Chinese writing system and the ancestor of modern Chinese.
1 code implementation • 29 Sep 2020 • Yusheng Su, Xu Han, Zhengyan Zhang, Peng Li, Zhiyuan Liu, Yankai Lin, Jie zhou, Maosong Sun
In this paper, we propose a novel framework named Coke to dynamically select contextual knowledge and embed knowledge context according to textual context for PLMs, which can avoid the effect of redundant and ambiguous knowledge in KGs that cannot match the input text.
no code implementations • 19 Sep 2020 • Zheni Zeng, Chaojun Xiao, Yuan YAO, Ruobing Xie, Zhiyuan Liu, Fen Lin, Leyu Lin, Maosong Sun
Recommender systems aim to provide item recommendations for users, and are usually faced with data sparsity problem (e. g., cold start) in real-world scenarios.
1 code implementation • ACL 2021 • Guoyang Zeng, Fanchao Qi, Qianrui Zhou, Tingji Zhang, Zixian Ma, Bairu Hou, Yuan Zang, Zhiyuan Liu, Maosong Sun
Textual adversarial attacking has received wide and increasing attention in recent years.
no code implementations • 19 Sep 2020 • Yuan Zang, Bairu Hou, Fanchao Qi, Zhiyuan Liu, Xiaojun Meng, Maosong Sun
Adversarial attacking aims to fool deep neural networks with adversarial examples.
1 code implementation • 12 Sep 2020 • Huimin Chen, Zeyu Zhu, Fanchao Qi, Yining Ye, Zhiyuan Liu, Maosong Sun, Jianbin Jin
Therefore, in this study, we take China as a specific and typical case and investigate its image with aspect-based sentiment analysis on a large-scale Twitter dataset.
Aspect-Based Sentiment Analysis Aspect-Based Sentiment Analysis (ABSA)
1 code implementation • 14 Jul 2020 • Xuancheng Huang, Jiacheng Zhang, Zhixing Tan, Derek F. Wong, Huanbo Luan, Jingfang Xu, Maosong Sun, Yang Liu
System combination is an important technique for combining the hypotheses of different machine translation systems to improve translation performance.
no code implementations • ACL 2020 • Xu Han, Yi Dai, Tianyu Gao, Yankai Lin, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Continual relation learning aims to continually train a model on new data to learn incessantly emerging novel relations while avoiding catastrophically forgetting old relations.
2 code implementations • 2020 • Cheng Yang, Maosong Sun, Zhiyuan Liu, Cunchao Tu
Many Network Representation Learning (NRL) methods have been proposed to learn vector representations for vertices in a network recently.
1 code implementation • Findings (ACL) 2021 • Jie Zhou, Shengding Hu, Xin Lv, Cheng Yang, Zhiyuan Liu, Wei Xu, Jie Jiang, Juanzi Li, Maosong Sun
Based on the datasets, we propose novel tasks such as multi-hop knowledge abstraction (MKA), multi-hop knowledge concretization (MKC) and then design a comprehensive benchmark.
2 code implementations • ACL 2020 • Haoxi Zhong, Chaojun Xiao, Cunchao Tu, Tianyang Zhang, Zhiyuan Liu, Maosong Sun
Legal Artificial Intelligence (LegalAI) focuses on applying the technology of artificial intelligence, especially natural language processing, to benefit tasks in the legal domain.
1 code implementation • EMNLP 2020 • Yuxian Gu, Zhengyan Zhang, Xiaozhi Wang, Zhiyuan Liu, Maosong Sun
In this stage, the model is trained by masked language modeling on in-domain unsupervised data to learn domain-specific patterns and we propose a novel selective masking strategy to learn task-specific patterns.
2 code implementations • EMNLP 2020 • Deming Ye, Yankai Lin, Jiaju Du, Zheng-Hao Liu, Peng Li, Maosong Sun, Zhiyuan Liu
Language representation models such as BERT could effectively capture contextual semantic information from plain text, and have been proved to achieve promising results in lots of downstream NLP tasks with appropriate fine-tuning.
Ranked #31 on Relation Extraction on DocRED
no code implementations • Asian Chapter of the Association for Computational Linguistics 2020 • Xu Han, Tianyu Gao, Yankai Lin, Hao Peng, Yaoliang Yang, Chaojun Xiao, Zhiyuan Liu, Peng Li, Maosong Sun, Jie zhou
Relational facts are an important component of human knowledge, which are hidden in vast amounts of text.
no code implementations • 13 Mar 2020 • Xiaoyuan Yi, Ruoyu Li, Cheng Yang, Wenhao Li, Maosong Sun
Though recent neural models make prominent progress in some criteria of poetry quality, generated poems still suffer from the problem of poor diversity.
no code implementations • LREC 2020 • Jinyi Hu, Maosong Sun
In this paper, we propose a GPT-2 based uniformed framework for generating major types of Chinese classical poems.
1 code implementation • 16 Jan 2020 • Jiaju Du, Fanchao Qi, Maosong Sun, Zhiyuan Liu
We find that sememes of each word are usually semantically matched to different words in its dictionary definition, and we name this matching relationship local semantic correspondence.
1 code implementation • 18 Dec 2019 • Lei Zhang, Fanchao Qi, Zhiyuan Liu, Yasheng Wang, Qun Liu, Maosong Sun
A reverse dictionary takes the description of a target word as input and outputs the target word together with other words that match the description.
no code implementations • 5 Dec 2019 • Gang Chen, Yang Liu, Huanbo Luan, Meng Zhang, Qun Liu, Maosong Sun
While the use of neural networks has proven effective in improving story generation, how to learn to generate an explainable high-level plot still remains a major challenge.
3 code implementations • 4 Dec 2019 • Fanchao Qi, Liang Chang, Maosong Sun, Sicong Ouyang, Zhiyuan Liu
We first build a dataset serving as the seed of the multilingual sememe KB.