Publications (22)
FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts
Yichen Gong, Delong Ran, Jinyuan Liu +5
Large Vision-Language Models (LVLMs) signify a groundbreaking paradigm shift within the Artificial Intelligence (AI) community, extending beyond the capabilities of Large Language…
Can VLMs Detect and Localize Fine-Grained AI-Edited Images?
Zhen Sun, Ziyi Zhang, Zeren Luo +10
Fine-grained detection and localization of localized image edits is crucial for assessing content authenticity, especially as modern diffusion models and image editors can produce…
On Evaluating The Performance of Watermarked Machine-Generated Texts Under Adversarial Attacks
Zesen Liu, Tianshuo Cong, Xinlei He +1
Large Language Models (LLMs) excel in various applications, including text generation and complex tasks. However, the misuse of LLMs raises concerns about the authenticity and ethi…
Beyond the Tip of Efficiency: Uncovering the Submerged Threats of Jailbreak Attacks in Small Language Models
Sibo Yi, Tianshuo Cong, Xinlei He +2
Small language models (SLMs) have become increasingly prominent in the deployment on edge devices due to their high efficiency and low computational cost. While researchers continu…
The Benchmark Illusion: Pruned LLMs Can Pass Multiple Choice but Fail to Answer
Rui Wen, Lu Sun, Jiayang Liu +3
Compressing large language models reduces memory use and inference cost, but it can also create failures that standard benchmarks miss. A pruned model may still perform well on mul…
LoRA-Leak: Membership Inference Attacks Against LoRA Fine-tuned Language Models
Delong Ran, Xinlei He, Tianshuo Cong +3
Language Models (LMs) typically adhere to a "pre-training and fine-tuning" paradigm, where a universal pre-trained model can be fine-tuned to cater to various specialized domains.…
FakeI2V-Bench: Benchmarking the Applicability of Image-level Deepfake Detectors for Deepfake Video Detection
Pei Li, Sihan Chen, Delong Ran +1
Recent advances in video generation models have significantly intensified the deepfake threat, yet the current deepfake video detection benchmarks remain underdeveloped. In particu…
Evaluating Implicit Regulatory Compliance in LLM Tool Invocation via Logic-Guided Synthesis
Da Song, Yuheng Huang, Boqi Chen +4
The integration of large language models (LLMs) into autonomous agents has enabled complex tool use, yet in high-stakes domains, these systems must strictly adhere to regulatory st…
PEFTGuard: Detecting Backdoor Attacks Against Parameter-Efficient Fine-Tuning
Zhen Sun, Tianshuo Cong, Yule Liu +5
Fine-tuning is an essential process to improve the performance of Large Language Models (LLMs) in specific domains, with Parameter-Efficient Fine-Tuning (PEFT) gaining popularity d…
SSLGuard: A Watermarking Scheme for Self-supervised Learning Pre-trained Encoders
Tianshuo Cong, Xinlei He, Yang Zhang
Self-supervised learning is an emerging machine learning paradigm. Compared to supervised learning which leverages high-quality labeled datasets, self-supervised learning relies on…
CL-Attack: Textual Backdoor Attacks via Cross-Lingual Triggers
Jingyi Zheng, Tianyi Hu, Tianshuo Cong +1
Backdoor attacks significantly compromise the security of large language models by triggering them to output specific and controlled content. Currently, triggers for textual backdo…
Have You Merged My Model? On The Robustness of Large Language Model IP Protection Methods Against Model Merging
Tianshuo Cong, Delong Ran, Zesen Liu +6
Model merging is a promising lightweight model empowerment technique that does not rely on expensive computing devices (e.g., GPUs) or require the collection of specific training d…
Robust Watermarks Meet Backdoored Models: Evading Diffusion Semantic Watermarks via Stealthy Backdoor
Jinyuan Liu, Tianshuo Cong, Pei Li +4
Although semantic watermarking is considered a promising safeguard for images generated by Latent Diffusion Models (LDMs), the reliance of the watermark detection pipeline on neura…
SoK: Benchmarking Poisoning Attacks and Defenses in Federated Learning
Heyi Zhang, Yule Liu, Xinlei He +3
Federated learning (FL) enables collaborative model training while preserving data privacy, but its decentralized nature exposes it to client-side data poisoning attacks (DPAs) and…
Watermarking LLM-Generated Datasets in Downstream Tasks
Yugeng Liu, Tianshuo Cong, Michael Backes +2
Large Language Models (LLMs) have experienced rapid advancements, with applications spanning a wide range of fields, including sentiment classification, review generation, and ques…
JailbreakEval: An Integrated Toolkit for Evaluating Jailbreak Attempts Against Large Language Models
Delong Ran, Jinyuan Liu, Yichen Gong +4
Jailbreak attacks induce Large Language Models (LLMs) to generate harmful responses, posing severe misuse threats. Though research on jailbreak attacks and defenses is emerging, th…
Robustness Over Time: Understanding Adversarial Examples' Effectiveness on Longitudinal Versions of Large Language Models
Yugeng Liu, Tianshuo Cong, Zhengyu Zhao +3
Large Language Models (LLMs) undergo continuous updates to improve user experience. However, prior research on the security and safety implications of LLMs has primarily focused on…
Test-Time Poisoning Attacks Against Test-Time Adaptation Models
Tianshuo Cong, Xinlei He, Yun Shen +1
Deploying machine learning (ML) models in the wild is challenging as it suffers from distribution shifts, where the model trained on an original domain cannot generalize well to un…
Cryptanalysis of LDPC-Based Pseudorandom Error-Correcting Codes
Tianrui Wang, Anyu Wang, Tianshuo Cong +3
Pseudorandom error-correcting codes (PRCs), a novel cryptographic primitive recently proposed at CRYPTO 2024, are primarily applied in undetectable watermarking schemes for large g…
Auditing Data Membership in Reinforcement Learning With Verifiable Rewards
Yule Liu, Heyi Zhang, Jinyi Zheng +6
Reinforcement Learning with Verifiable Rewards (RLVR) has become a core training stage in recent large language models (LLMs). Its reliance on non-public, high-value prompt sets ra…
Jailbreak Attacks and Defenses Against Large Language Models: A Survey
Sibo Yi, Yule Liu, Zhen Sun +5
Large Language Models (LLMs) have performed exceptionally in various text-generative tasks, including question answering, translation, code completion, etc. However, the over-assis…
From Defender to Devil? Unintended Risk Interactions Induced by LLM Defenses
Xiangtao Meng, Tianshuo Cong, Li Wang +4
Large Language Models (LLMs) have shown remarkable performance across various applications, but their deployment in real-world settings faces several risks, including jailbreak att…