Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Yu Liang

Yu Liang

4 indexed papers

Recent (6 mo)
4
With code
0
Influential cites
0
Benchmarked
0

Publications per year

4
26

Top categories

AI×2Crypto×2ML×1

Frequent co-authors

Yibing Liu2×
Zhen Chen1×
Weihao Xie1×
Peilin Chen1×
Shiqi Wang1×
Guoxin Ma1×

Research Timeline

2026
RLSpoofer: A Lightweight Evaluator for LLM Watermark Spoofing Resilience

The paper introduces RLSpoofer, a lightweight, black-box reinforcement learning attack that demonstrates the fragile resilience of current LLM watermarking schemes by achieving a high spoofing success rate with minimal training data.

Thinking as Compression: Your Reasoning Model is Secretly a Context Compressor

The paper introduces Thinking as Compression (TaC), a novel paradigm showing that the inherent reasoning process of a large language model can naturally compress long context inputs, outperforming dedicated compression methods.

Revisiting ML Training under Fully Homomorphic Encryption: Convergence Guarantees, Differential Privacy, and Efficient Algorithms

The paper provides the first theoretical convergence analysis for machine learning training under fully homomorphic encryption combined with differential privacy, improving efficiency and scalability.

RAISE: RAG Design as an Architecture Search Problem

The paper proposes formulating RAG design as an architecture search problem and introduces RAISE, a comprehensive framework and benchmark for systematically optimizing RAG hyperparameters.

Highlighted terms show continued research focus across papers

Papers

cs.AIRecentMay 28, 2026

RAISE: RAG Design as an Architecture Search Problem

Zhen Chen, Yibing Liu, Weihao Xie, Yu Liang +2 more

The paper proposes formulating RAG design as an architecture search problem and introduces RAISE, a comprehensive framework and benchmark for systematically optimizing RAG hyperparameters.

View →
cs.AIRecentMay 27, 2026

Thinking as Compression: Your Reasoning Model is Secretly a Context Compressor

Guoxin Ma, Yibing Liu, Chengzhengxu Li, Yu Liang +6 more

The paper introduces Thinking as Compression (TaC), a novel paradigm showing that the inherent reasoning process of a large language model can naturally compress long context inputs, outperforming ded…

View →
cs.LGcs.CRRecentMay 27, 2026

Revisiting ML Training under Fully Homomorphic Encryption: Convergence Guarantees, Differential Privacy, and Efficient Algorithms

Yvonne Zhou, Mingyu Liang, Ivan Brugere, Danial Dervovic +4 more

The paper provides the first theoretical convergence analysis for machine learning training under fully homomorphic encryption combined with differential privacy, improving efficiency and scalability.

View →
cs.CRRecentApr 13, 2026

RLSpoofer: A Lightweight Evaluator for LLM Watermark Spoofing Resilience

Hanbo Huang, Xuan Gong, Yiran Zhang, Hao Zheng +1 more

The paper introduces RLSpoofer, a lightweight, black-box reinforcement learning attack that demonstrates the fragile resilience of current LLM watermarking schemes by achieving a high spoofing success…

View →