GD
Gelei Deng
cs.AIcs.CRcs.SEcs.CLcs.LGcs.CVcs.CYcs.SDcs.SIeess.AS
On Valency
published · living versionsW_9j87nudb·v1 · currentpublished
Silent and Costly: Characterizing and Detecting Prompt-Caching Mis-Integration in LLM Agents
with Ying Zhang, Ruoqi Guo, Yi Liu, Yuekang Li +2
1 version
Preprints & journals
64 papers in the corpus · 2021–2026One Skill Too Many: How Co-Installed Skills Conflict in Coding Agents2610.11647v2 · Chaoliang Yan, Zihao Xu, Yuekang Li et al.2026 · 0 citationsarXiv
AI Slop and Hallucinations in Vulnerability Assessment: A Survey on Reasoning Failures and Trustworthy Mitigation2608.25667v1 · Junchen Ding, Jialiang Dong, Yichen Zhu et al.2026 · 0 citationsarXiv
MaliciousSkillBench: A Comprehensive Benchmark for Malicious Agent Skill Detection2608.19901v1 · Yue Wang, Yi Liu, Gelei Deng et al.2026 · 0 citationsarXiv
Auditing Chinese Web-scale Corpora via Sampled BPE Token Statistics2608.10678v2 · Qingjie Zhang, Ziqi Tang, Jie Zhang et al.2026 · 0 citationsarXiv
When Claws Remember but Do Not Tell: Stealthy Memory Injection in Persistent Personal Agents2607.05189v1 · Yechao Zhang, Shiqian Zhao, Jiawen Zhang et al.2026 · 0 citationsarXiv
Rethinking Complexity Metrics for LLM-Integrated Applications: Beyond Source Code2607.01903v1 · Zihao Xu, Yuekang Li, Gelei Deng et al.2026 · 0 citationsarXiv
A Survey of LLM-Driven Penetration Testing: Taxonomy, Co-Evolution, and Open Challenges2607.02605v1 · Zheyuan He, Jiaxun Dong, Zihao Li et al.2026 · 0 citationsarXiv
MIRAGE: Context-Aware Prompt Injection against Mobile GUI Agents via User-Generated Content2605.28116v1 · Ruoqi Guo, Yi Liu, Gelei Deng et al.2026 · 0 citationsarXiv
SNARE: Adaptive Scenario Synthesis for Eliciting Overeager Behavior in Coding Agents2605.28122v1 · Yubin Qu, Yi Liu, Gelei Deng et al.2026 · 0 citationsarXiv
Turning Bias into Bugs: Bandit-Guided Style Manipulation Attacks on LLM Judges2605.26156v2 · Xianglin Yang, Bryan Hooi, Gelei Deng et al.2026 · 0 citationsarXiv
Overeager Coding Agents: Measuring Out-of-Scope Actions on Benign Tasks2605.18583v1 · Yubin Qu, Ying Zhang, Yanjun Zhang et al.2026 · 0 citationsarXiv
Membership Inference Attacks Against Video Large Language Models2604.27002v1 · Wei Song, Yuxin Cao, Ziqi Ding et al.2026 · 0 citationsarXiv
How Your Credentials Are Leaked by LLM Agent Skills: An Empirical Study2604.03070v2 · Zhihao Chen, Ying Zhang, Yi Liu et al.2026 · 0 citationsarXiv
Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems2604.03081v1 · Yubin Qu, Yi Liu, Tongcheng Geng et al.2026 · 0 citationsarXiv
AutoEG: Exploiting Known Third-Party Vulnerabilities in Black-Box Web Applications2604.00704v1 · Ruozhao Yang, Mingfei Cheng, Gelei Deng et al.2026 · 0 citationsarXiv
Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution2603.23064v3 · Yechao Zhang, Shiqian Zhao, Jie Zhang et al.2026 · 0 citationsarXiv
"Are You Sure?": An Empirical Study of Human Perception Vulnerability in LLM-Driven Agentic Systems2602.21127v1 · Xinfeng Li, Shenyu Dai, Kelong Zheng et al.2026 · 0 citationsarXiv
Do LLMs and VLMs Share Neurons for Inference? Evidence and Mechanisms of Cross-Modal Transfer2602.19058v1 · Chenhang Cui, An Zhang, Yuxin Chen et al.2026 · 0 citationsarXiv
What Makes a Good LLM Agent for Real-world Penetration Testing?2602.17622v1 · Gelei Deng, Yi Liu, Yuekang Li et al.2026 · 0 citationsarXiv
"Do Not Mention This to the User": Detecting and Understanding Malicious Agent Skills in the Wild2602.06547v4 · Yi Liu, Zhihao Chen, Yanjun Zhang et al.2026 · 0 citationsarXiv
Risky-Bench: Evaluating Agentic Safety Risks in Real-World Life-Assist Deployments2602.03100v2 · Jingnan Zheng, Yanzhen Luo, Jingjun Xu et al.2026 · 0 citationsarXiv
Self-Guard: Defending Large Reasoning Models via enhanced self-reflection2602.00707v1 · Jingnan Zheng, Jingjun Xu, Yanzhen Luo et al.2026 · 0 citationsarXiv
DECEIVE-AFC: Adversarial Claim Attacks against Search-Enabled LLM-based Fact-Checking Systems2602.02569v2 · Haoran Ou, Kangjie Chen, Gelei Deng et al.2026 · 0 citationsarXiv
Agent Skills in the Wild: An Empirical Study of Security Vulnerabilities at Scale2601.10338v1 · Yi Liu, Weizhe Wang, Ruitao Feng et al.2026 · 0 citationsarXiv
Robust CAPTCHA Using Audio Illusions in the Era of Large Language Models: from Evaluation to Advances2601.08516v1 · Ziqi Ding, Yunfeng Wan, Wei Song et al.2026 · 0 citationsarXiv
PentestEval: Benchmarking LLM-based Penetration Testing with Modular and Stage-Level Design2512.14233v1 · Ruozhao Yang, Mingfei Cheng, Gelei Deng et al.2025 · 0 citationsarXiv
When Search Goes Wrong: Red-Teaming Web-Augmented Large Language Models2510.09689v3 · Haoran Ou, Kangjie Chen, Xingshuo Han et al.2025 · 0 citationsarXiv
Visible Yet Unreadable: A Systematic Blind Spot of Vision Language Models Across Writing Systems2509.06996v5 · Jie Zhang, Ting Xu, Gelei Deng et al.2025 · 0 citationsarXiv
When Audio and Text Disagree: Revealing Text Bias in Large Audio-Language Models2508.15407v1 · Cheng Wang, Gelei Deng, Xianglin Yang et al.2025 · 4 citationsarXiv
CognixShield: PoV-Guided Vulnerable API Usage Detection in Large Codebases via LLMs2506.17798v3 · Quanzhi Fu, Wang Lingxiang, Wenjia Song et al.2025 · 0 citationsarXiv
RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards2506.07736v3 · Jingnan Zheng, Xiangtian Ji, Yijun Lu et al.2025 · 0 citations39th Conference on Neural Information Processing Systems (NeurIPS 2025). 39th Conference on Neural Information Processing Systems (NeurIPS 2025). 39th Conference on Neural Information Processing Systems (NeurIPS 2025)
IRCopilot: Automated Incident Response with Large Language Models2505.20945v3 · Xihuan Lin, Jie Zhang, Gelei Deng et al.2025 · 1 citationarXiv
Beyond Retrieval: Improving Evidence Quality for LLM-based Multimodal Fact-Checking2505.03135v2 · Haoran Ou, Gelei Deng, Xingshuo Han et al.2025 · 0 citationsarXiv
A Rusty Link in the AI Supply Chain: Detecting Evil Configurations in Model Repositories2505.01067v1 · Ziqi Ding, Qian Fu, Junchen Ding et al.2025 · 2 citationsarXiv
IllusionCAPTCHA: A CAPTCHA based on Visual Illusion2502.05461v2 · Ziqi Ding, Gelei Deng, Yi Liu et al.2025 · 11 citationsarXiv
Enhancing Model Defense Against Jailbreaks with Proactive Safety Reasoning2501.19180v2 · Xianglin Yang, Gelei Deng, Jieming Shi et al.2025 · 0 citationsarXiv
Safe + Safe = Unsafe? Exploring How Safe Images Can Be Exploited to Jailbreak Large Vision-Language Models2411.11496v3 · Chenhang Cui, Gelei Deng, An Zhang et al.2024 · 0 citationsarXiv
Fine-Grained Verifiers: Preference Modeling as Next-token Prediction in Vision-Language Alignment2410.14148v4 · Chenhang Cui, An Zhang, Yiyang Zhou et al.2024 · 0 citationsarXiv
GenderCARE: A Comprehensive Framework for Assessing and Reducing Gender Bias in Large Language Models2408.12494v2 · Kunsheng Tang, Wenbo Zhou, Jie Zhang et al.2024 · 14 citationsarXiv
Efficient Detection of Toxic Prompts in Large Language Models2408.11727v3 · Yi Liu, Junzhe Yu, Huijia Sun et al.2024 · 8 citationsarXiv
Image-Based Geolocation Using Large Vision-Language Models2408.09474v1 · Yi Liu, Junchen Ding, Gelei Deng et al.2024 · 0 citationsarXiv
SPOLRE: Semantic Preserving Object Layout Reconstruction for Image Captioning System Testing2407.18512v1 · Yi Liu, Guanyu Wang, Xinyi Zheng et al.2024 · 0 citationsarXiv
Continuous Embedding Attacks via Clipped Inputs in Jailbreaking Large Language Models2407.13796v1 · Zihao Xu, Yi Liu, Gelei Deng et al.2024 · 0 citationsarXiv
Source Code Summarization in the Era of Large Language Models2407.07959v2 · Weisong Sun, Yun Miao, Yuekang Li et al.2024 · 64 citationsarXiv
GeneRAG: Enhancing Large Language Models with Gene-Related Task by Retrieval-Augmented Generation10.1101/2024.06.24.600176v1 · Lin, X., Deng, G., Li, Y. et al.2024 · 7 citationsbioRxiv
Uncovering Logit Suppression Vulnerabilities in LLM Safety Alignment2405.13068v4 · Yuxi Li, Yi Liu, Yuekang Li et al.2024 · 4 citationsDe Marsico, M., et al. Pattern Recognition. ICPR 2026. Lecture Notes in Computer Science, vol 16813. Springer, Cham
Oedipus: LLM-enchanced Reasoning CAPTCHA Solver2405.07496v1 · Gelei Deng, Haoran Ou, Yi Liu et al.2024 · 3 citationsarXiv
Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection2404.09894v3 · Yuxi Li, Yi Liu, Gelei Deng et al.2024 · 13 citationsarXiv
MeTMaP: Metamorphic Testing for Detecting False Vector Matching Problems in LLM Augmented Generation2402.14480v1 · Guanyu Wang, Yuekang Li, Yi Liu et al.2024 · 13 citationsarXiv
A Comprehensive Study of Jailbreak Attack versus Defense for Large Language Models2402.13457v2 · Zihao Xu, Yi Liu, Gelei Deng et al.2024 · 69 citationsarXiv
Career total: 84 works. 64 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.