3 citations · 7 across the 41 of their papers we have counts for
46 papers
Probabilistic Multi-Source Data Selection via Learnable Categorical Priors: Variance Reduction and Guaranteed Derandomization
Shizhan Cai, Qixin Zhang, Siyuan Ma +3
Bilevel data selection is a fundamental task with numerous applications in machine learning and statistics. A common strategy for tackling this problem is to relax the original bil…
TrojanWorld: Backdooring World-Model Agents via Imagination Steering
Wenkai Huang, Siyuan Liang, Gaolei Li +4
World models increasingly serve as the predictive core of model-based reinforcement learning agents, enabling them to simulate future dynamics and reason over imagined trajectories…
CIVA: Critic-Induced Value-Subspace Attacks on Visual World-Model Agents
Jiancheng Wang, Mingli Zhu, Tong Zhang +4
Visual world-model agents such as DreamerV3 act through a recurrent latent state rather than a single observation, which weakens frame-wise observation attacks and makes their pert…
Breadcrumbing Search Agents
Xuebin Li, Hanqing Zhao, Siyuan Liang +4
LLM-based search agents are widely used for information-seeking tasks, but their reliance on external tool returns introduces a critical security risk: web content retrieved during…
SafeCA: Safe Cross-Attention Localization and Regulation for Text-to-Video Jailbreak Defense
Siyuan Liang, Yupeng Qiu, Junfeng Fang +3
Text-to-Video (T2V) generative models are vulnerable to jailbreak attacks in real-world deployment, leading them to produce harmful or inappropriate content. Existing defense appro…
SoK: Intent-Oriented Systematization of Multi-Turn LLM Jailbreaks
Siyuan Li, Aodu Wulianghai, Zehao Liu +8
Large Language Models (LLMs) are increasingly deployed in interactive settings, where user intent commonly unfolds through multi-turn dialogue. Multi-turn jailbreaks exploit this p…
Benchmarking the Robustness of Autonomous Driving to Environmental Illusions: A Lane Perception Perspective
Tianyuan Zhang, Xianglong Liu, Aishan Liu +6
Environmental illusions (eg., shadows, reflections, and tire marks) are naturally existing yet overlooked phenomena in real-world driving environments. They can disturb visual perc…
Cert-LAS: Toward Certified Model Ownership Verification for Text-to-Image Diffusion Models via Layer-Adaptive Smoothing
Leyi Qi, Yiming Li, Siyuan Liang +2
Large-scale text-to-image (T2I) diffusion models have enabled unprecedented creative applications, but their unauthorized use has raised serious intellectual property concerns, mak…
TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting
Quang Duc Nguyen, Siyuan Liang, Yiming Li +2
Time Series Forecasting (TSF) is highly vulnerable to backdoor attacks, yet effective defenses remain underexplored due to challenges arising from data entanglement and shifts in t…
Who Generated This 3D Asset? Learning Source Attribution for Generative 3D Models
Sihan Ma, Siyuan Liang, Dacheng Tao
Generative 3D models are deployed in gaming, robotics, and immersive creation, making source attribution critical: given a 3D asset, can we identify whether and which generative mo…
WMAttack: Automated Attack Search for Adversarial Evaluation of World-Model Agents
Zhixiang Guo, Siyuan Liang, Shi Fu +4
Despite the growing use of world models as decision-making agents, their adversarial robustness remains underexplored due to the lack of dedicated automated evaluation methods. A k…
A Survey of Large Audio Language Models: Generalization, Trustworthiness, and Outlook
Kaiwen Luo, Zhenhong Zhou, Leyan Wang +32
Advances in Large Language Models (LLMs) have paved the way for Multimodal Large Language Models (MLLMs). Among these, Large Audio Language Models (LALMs) are essential for realizi…
CtrlAttack: A Unified Attack on World-Model Control in Diffusion Models
Shuhan Xu, Siyuan Liang, Hongling Zheng +4
Diffusion-based image-to-video (I2V) models increasingly exhibit world-model-like properties by implicitly capturing temporal dynamics. However, existing studies have mainly focuse…
When World Models Dream Wrong: Physical-Conditioned Adversarial Attacks against World Models
Zhixiang Guo, Siyuan Liang, Andras Balogh +4
Generative world models (WMs) are increasingly used to synthesize controllable, sensor-conditioned driving videos, yet their reliance on physical priors exposes novel attack surfac…
Towards Reliable Medical LLMs: Benchmarking and Enhancing Confidence Estimation of Large Language Models in Medical Consultation
Zhiyao Ren, Yibing Zhan, Siyuan Liang +3
Large-scale language models (LLMs) often offer clinical judgments based on incomplete information, increasing the risk of misdiagnosis. Existing studies have primarily evaluated co…
Review of Hallucination Understanding in Large Language and Vision Models
Zhengyi Ho, Siyuan Liang, Dacheng Tao
The widespread adoption of large language and vision models in real-world applications has made urgent the need to address hallucinations -- instances where models produce incorrec…
Detoxifying Large Language Models via Autoregressive Reward Guided Representation Editing
Yisong Xiao, Aishan Liu, Siyuan Liang +3
Large Language Models (LLMs) have demonstrated impressive performance across various tasks, yet they remain vulnerable to generating toxic content, necessitating detoxification str…
Universal Camouflage Attack on Vision-Language Models for Autonomous Driving
Dehong Kong, Sifan Yu, Siyuan Liang +4
Visual language modeling for automated driving is emerging as a promising research direction with substantial improvements in multimodal reasoning capabilities. Despite its advance…
Adversarial Generation and Collaborative Evolution of Safety-Critical Scenarios for Autonomous Vehicles
Jiangfan Liu, Yongkang Guo, Fangzhi Zhong +7
The generation of safety-critical scenarios in simulation has become increasingly crucial for safety evaluation in autonomous vehicles prior to road deployment in society. However,…
Bench2ADVLM: A Closed-Loop Benchmark for Vision-language Models in Autonomous Driving
Tianyuan Zhang, Ting Jin, Lu Wang +5
Vision-Language Models (VLMs) have recently emerged as a promising paradigm in autonomous driving (AD). However, current performance evaluation protocols for VLM-based AD systems (…
PromptSafe: Gated Prompt Tuning for Safe Text-to-Image Generation
Zonglei Jing, Xiao Yang, Xiaoqian Li +4
Text-to-image (T2I) models have demonstrated remarkable generative capabilities but remain vulnerable to producing not-safe-for-work (NSFW) content, such as violent or explicit ima…
MetAdv: A Unified and Interactive Adversarial Testing Platform for Autonomous Driving
Aishan Liu, Jiakai Wang, Tianyuan Zhang +6
Evaluating and ensuring the adversarial robustness of autonomous driving (AD) systems is a critical and unresolved challenge. This paper introduces MetAdv, a novel adversarial test…
Pushing the Limits of Safety: A Technical Report on the ATLAS Challenge 2025
Zonghao Ying, Siyang Wu, Run Hao +44
Multimodal Large Language Models (MLLMs) have enabled transformative advancements across diverse applications but remain susceptible to safety threats, especially jailbreak attacks…
ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks
Zhiyao Ren, Siyuan Liang, Aishan Liu +1
In-context learning (ICL) has demonstrated remarkable success in large language models (LLMs) due to its adaptability and parameter-free nature. However, it also introduces a criti…