SYSU | ASsured data Centric Learning Lab

Rigorous Science for Trustworthy Data & AI

We study the security, privacy, and reliability of AI — spanning large language models, federated learning, time series intelligence, and modern data systems — to build rigorous theory, practical algorithms, and deployable systems for trustworthy AI and industrial intelligence.

Meet the Team →Publications →
LLM SecurityLLM PrivacyDifferential PrivacyFederated LearningTime Series AIData ValuationMachine UnlearningData Governance
ASCllTrustworthyLLMFederatedLearningDifferentialPrivacyDataValuationTime SeriesIntelligenceRAG &Agent AI

Research Focus

What We Work On

🛡

LLM Security & Alignment

Backdoor defense, jailbreak detection, adversarial robustness, quantization-aware safety.

🔐

Privacy & Federated Learning

Differential privacy, secure MPC, personalized FL, machine unlearning.

📈

Time Series Intelligence

Anomaly detection, predictive maintenance, time series foundation models.

💎

Data Markets & Valuation

Fair data pricing, data rights, end-to-end marketplace infrastructure.

View all research areas →

Lab at a Glance

120+

Peer-reviewed Publications

50+

CCF-A / Top-tier Papers

6+

Best Paper Awards

5+

Global Collaborations

Lab Life

Daily Activities

Outings, seminars, celebrations — candid moments from the lab.

Open gallery →

Principal Investigators

Dan Li

Dan Li (李丹)

Associate Professor · CPS, Time Series, Industrial AI

MAD-GAN · 1660+ citations

Jian Lou

Jian Lou (娄坚)

Associate Professor · LLM Privacy, Data Valuation

CCS 2024 Distinguished Paper

Latest Work

Recent Publications

ICLR 2026

Rating Quality of Diverse Time Series Data by Meta-learning from LLM Judgment

DASFAA 2026

Lightweight Time Series Data Valuation on Time Series Foundation Models via In-Context Finetuning

ACM CCSCCF-A

PreferCare: Preference Dataset Copyright Protection in LLM Alignment by Watermark Injection and Verification

USENIX SecCCF-A

Activation Approximations Can Incur Safety Vulnerabilities in Aligned LLMs: Comprehensive Analysis and Defense

ICMLCCF-A

PoisonedEye: Knowledge Poisoning Attack on RAG-based Large Vision-Language Models

View all publications →

Join Us

We are actively recruiting PhD students, Master's students, postdocs, and undergraduate research interns. We provide hands-on mentorship, international collaboration opportunities, and a supportive team culture.

Learn More →Meet the Team →