Lei Sha
Scholar

Lei Sha

Google Scholar ID: EbZ_P6gAAAAJ
Prof@Beihang University, Prof@ZGC Lab, Oxtium AI, University of Oxford
NLPML
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
2,052
 
H-index
22
 
i10-index
30
 
Publications
20
 
Co-authors
28
list available
Contact
GitHubOpen ↗LinkedInOpen ↗
Publications
15 items
Be Your Own Red Teamer: Safety Alignment via Self-Play and Reflective Experience Replay
2026
Cited
0
BBox DocVQA: A Large Scale Bounding Box Grounded Dataset for Enhancing Reasoning in Document Visual Question Answer
2025
Cited
0
Benchmarking Multi-Step Legal Reasoning and Analyzing Chain-of-Thought Effects in Large Language Models
2025
Cited
0
Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment
2025
Cited
0
HauntAttack: When Attack Follows Reasoning as a Shadow
2025
Cited
0
Towards Harmonized Uncertainty Estimation for Large Language Models
2025
Cited
0
AISafetyLab: A Comprehensive Framework for AI Safety Evaluation and Improvement
2025
Cited
0
Be a Multitude to Itself: A Prompt Evolution Framework for Red Teaming
Conference on Empirical Methods in Natural Language Processing · 2025
Cited
0
Resume (English only)
Co-authors
28 total
Zhifang SUI
Zhifang SUI
Peking University
Baobao CHANG
Baobao CHANG
Peking University
Sujian LI
Sujian LI
Peking Univ.
Thomas Lukasiewicz
Thomas Lukasiewicz
Vienna University of Technology, Austria; University of Oxford, UK
Tianyu Liu
Tianyu Liu
Alibaba
Co-author 6
Co-author 6
Co-author 7
Co-author 7
Hao Li
Hao Li
Beihang University