Yue Zhou

Yue Zhou, Henry Peng Zou, Barbara Di Eugenio et al.

We find that language models have difficulties generating fallacious and deceptive reasoning. When asked to generate deceptive outputs, language models tend to leak honest counterparts but believe them to be false. Exploiting this deficiency, we propose a jailbreak attack method that elicits an aligned language model for malicious output. Specifically, we query the model to generate a fallacious yet deceptively real procedure for the harmful behavior. Since a fallacious procedure is generally considered fake and thus harmless by LLMs, it helps bypass the safeguard mechanism. Yet the output is factually harmful since the LLM cannot fabricate fallacious solutions but proposes truthful ones. We evaluate our approach over five safety-aligned large language models, comparing four previous jailbreak methods, and show that our approach achieves competitive performance with more harmful outputs. We believe the findings could be extended beyond model safety, such as self-verification and hallucination.

Yue Zhou, Chenlu Guo, Xu Wang et al.

Large models, encompassing large language and diffusion models, have shown exceptional promise in approximating human-level intelligence, garnering significant interest from both academic and industrial spheres. However, the training of these large models necessitates vast quantities of high-quality data, and with continuous updates to these models, the existing reservoir of high-quality data may soon be depleted. This challenge has catalyzed a surge in research focused on data augmentation methods. Leveraging large models, these data augmentation techniques have outperformed traditional approaches. This paper offers an exhaustive review of large model-driven data augmentation methods, adopting a comprehensive perspective. We begin by establishing a classification of relevant studies into three main categories: image augmentation, text augmentation, and paired data augmentation. Following this, we delve into various data post-processing techniques pertinent to large model-based data augmentation. Our discussion then expands to encompass the array of applications for these data augmentation methods within natural language processing, computer vision, and audio signal processing. We proceed to evaluate the successes and limitations of large model-based data augmentation across different scenarios. Concluding our review, we highlight prospective challenges and avenues for future exploration in the field of data augmentation. Our objective is to furnish researchers with critical insights, ultimately contributing to the advancement of more sophisticated large models. We consistently maintain the related open-source materials at: https://github.com/MLGroup-JLU/LLM-data-aug-survey.

Yue Zhou, Barbara Di Eugenio

Despite LLMs' explicit alignment against demographic stereotypes, they have been shown to exhibit biases under various social contexts. In this work, we find that LLMs exhibit concerning biases in how they associate solution veracity with demographics. Through experiments across five human value-aligned LLMs on mathematics, coding, commonsense, and writing problems, we reveal two forms of such veracity biases: Attribution Bias, where models disproportionately attribute correct solutions to certain demographic groups, and Evaluation Bias, where models' assessment of identical solutions varies based on perceived demographic authorship. Our results show pervasive biases: LLMs consistently attribute fewer correct solutions and more incorrect ones to African-American groups in math and coding, while Asian authorships are least preferred in writing evaluation. In additional studies, we show LLMs automatically assign racially stereotypical colors to demographic groups in visualization code, suggesting these biases are deeply embedded in models' reasoning processes. Our findings indicate that demographic bias extends beyond surface-level stereotypes and social context provocations, raising concerns about LLMs' deployment in educational and evaluation settings.

Nan Wang, Sid Chi-Kin Chau, Yue Zhou

Energy storage provides an effective way of shifting temporal energy demands and supplies, which enables significant cost reduction under time-of-use energy pricing plans. Despite its promising benefits, the cost of present energy storage remains expensive, presenting a major obstacle to practical deployment. A more viable solution to improve the cost-effectiveness is by sharing energy storage, such as community sharing, cloud energy storage and peer-to-peer sharing. However, revealing private energy demand data to an external energy storage operator may compromise user privacy, and is susceptible to data misuses and breaches. In this paper, we explore a novel approach to support energy storage sharing with privacy protection, based on privacy-preserving blockchain and secure multi-party computation. We present an integrated solution to enable privacy-preserving energy storage sharing, such that energy storage service scheduling and cost-sharing can be attained without the knowledge of individual users' demands. It also supports auditing and verification by the grid operator via blockchain. Furthermore, our privacy-preserving solution can safeguard against a dishonest majority of users, who may collude in cheating, without requiring a trusted third-party. We implemented our solution as a smart contract on real-world Ethereum blockchain platform, and provide empirical evaluation in this paper.