Threats to Pre-trained Language Models: Survey and Taxonomy

Guo, Shangwei; Chunlong, Xie,; Li, Jiwei; Lyu, Lingjuan; Zhang, Tianwei

doi:10.48550/arxiv.2202.06862

Cited by 4 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The scale of used data is much larger than traditional methods, but it's still limited. The pursuit Vision-Language Intelligence: Tasks, Representation Learning, and Large Models [38] 2022 arXiv MM DC, 19 A survey on vision transformer [39] 2022 TPAMI CV DC, 23 Transformers in vision: A survey [40] 2021 CSUR CV SC, 38 A Survey of Visual Transformers [41] 2021 arXiv CV DC, 21 Video Transformers: A Survey [42] 2022 arXiv CV DC, 24 Threats to Pre-trained Language Models: Survey and Taxonomy [43] 2022 arXiv NLP DC, 8 A survey on bias in deep NLP [44] 2021 AS NLP SC, 26…”

Section: Conventional Deep Learningmentioning

confidence: 99%

See 1 more Smart Citation

Large-scale Multi-Modal Pre-trained Models: A Comprehensive Survey

Wang¹,

Chen²,

Qian³

et al. 2023

Preprint

View full text Add to dashboard Cite

With the urgent demand for generalized deep models, many pre-trained big models are proposed, such as BERT, ViT, GPT, etc. Inspired by the success of these models in single domains (like computer vision and natural language processing), the multi-modal pre-trained big models have also drawn more and more attention in recent years. In this work, we give a comprehensive survey of these models and hope this paper could provide new insights and helps fresh researchers to track the most cuttingedge works. Specifically, we firstly introduce the background of multi-modal pre-training by reviewing the conventional deep learning, pre-training works in natural language process, computer vision, and speech. Then, we introduce the task definition, key challenges, and advantages of multi-modal pretraining models (MM-PTMs), and discuss the MM-PTMs with a focus on data, objectives, network architectures, and knowledge enhanced pre-training. After that, we introduce the downstream tasks used for the validation of large-scale MM-PTMs, including generative, classification, and regression tasks. We also give visualization and analysis of the model parameters and results on representative downstream tasks. Finally, we point out possible research directions for this topic that may benefit future works. In addition, we maintain a continuously updated paper list for large-scale pre-trained multi-modal big models: https://github.com/wangxiao5791509/MultiModal BigModels Survey.

show abstract

Section: Conventional Deep Learningmentioning

confidence: 99%

“…The large-scale pre-trained models [29,43,44,[53][54][55][56] first appeared in the NLP field. Their success is mainly attributed to self-supervised learning and network structures like Transformer [9].…”

Section: Pre-training In Naturalmentioning

confidence: 99%

Large-scale Multi-Modal Pre-trained Models: A Comprehensive Survey

Wang¹,

Chen²,

Qian³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Privacy Protection To address privacy risks in NLP models, various privacy-preserving methods have been proposed, which can be categorized into three main stages of application (Guo et al, 2022;Sousa and Kern, 2023): data processing stage, pre-training and/or fine-tuning stage, and post-processing stage. In the data processing stage, methods involve removing or replacing sensitive information in the original data (Liu et al, 2017;El Emam et al, 2009;Zhou et al, 2008;García-Pablos et al, 2020).…”

Section: Related Workmentioning

confidence: 99%

“…Remarkable progress has been made in large language models (LLMs) in recent years (Brown et al, 2020;Liu et al, 2021;Ouyang et al, 2022;Lee et al, 2023). However,despite this success, LLMs are confronted with privacy and security concerns in real-world applications (Guo et al, 2022;Brown et al, 2022;Li et al, 2023). The primary cause of privacy and security risks is the inherent nature of large pretrained language models.…”

Section: Introductionmentioning

confidence: 99%

DEPN: Detecting and Editing Privacy Neurons in Pretrained Language Models

Wu,

Li,

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Large language models pretrained on a huge amount of data capture rich knowledge and information in the training data. The ability of data memorization and regurgitation in pretrained language models, revealed in previous studies, brings the risk of data leakage. In order to effectively reduce these risks, we propose a framework DEPN to Detect and Edit Privacy Neurons in pretrained language models, partially inspired by knowledge neurons and model editing. In DEPN, we introduce a novel method, termed as privacy neuron detector, to locate neurons associated with private information, and then edit these detected privacy neurons by setting their activations to zero. Furthermore, we propose a privacy neuron aggregator dememorize private information in a batch processing manner. Experimental results show that our method can significantly and efficiently reduce the exposure of private data leakage without deteriorating the performance of the model. Additionally, we empirically demonstrate the relationship between model memorization and privacy neurons, from multiple perspectives, including model size, training time, prompts, privacy neuron distribution, illustrating the robustness of our approach.

show abstract

“…Language-based VLMs inherit the risks of the underlying LLMs and vision models, such as gender and racial biases when prompted with images [52]. Several surveys on the ethics of LLMs are available [52,53]. Some work studies the robustness of VLMs against both natural distribution shifts [54] and adversarial robustness [55].…”

Section: Responsible Ai Considerations Of Promptingmentioning

confidence: 99%

Are Vision Transformers Robust to Patch Perturbations?

Gu¹,

Tresp²,

Qin

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Prompt engineering is a technique that involves augmenting a large pre-trained model with task-specific hints, known as prompts, to adapt the model to new tasks. Prompts can be created manually as natural language instructions or generated automatically as either natural language instructions or vector representations. Prompt engineering enables the ability to perform predictions based solely on prompts without updating model parameters, and the easier application of large pre-trained models in real-world tasks. In past years, Prompt engineering has been well-studied in natural language processing. Recently, it has also been intensively studied in vision-language modeling. However, there is currently a lack of a systematic overview of prompt engineering on pre-trained vision-language models. This paper aims to provide a comprehensive survey of cutting-edge research in prompt engineering on three types of vision-language models: multimodal-to-text generation models (e.g., Flamingo), image-text matching models (e.g., CLIP), and text-to-image generation models (e.g., Stable Diffusion). For each type of model, a brief model summary, prompting methods, prompting-based applications, and the corresponding responsibility and integrity issues are summarized and discussed. Furthermore, the commonalities and differences between prompting on vision-language models, language models, and vision models are also discussed. The challenges, future directions, and research opportunities are summarized to foster future research on this topic.

show abstract

Threats to Pre-trained Language Models: Survey and Taxonomy

Cited by 4 publications

References 0 publications

Large-scale Multi-Modal Pre-trained Models: A Comprehensive Survey

Large-scale Multi-Modal Pre-trained Models: A Comprehensive Survey

DEPN: Detecting and Editing Privacy Neurons in Pretrained Language Models

Are Vision Transformers Robust to Patch Perturbations?

Contact Info

Product

Resources

About