Zhenhong Zhou
Scholar

Zhenhong Zhou

Google Scholar ID: 6TuPwzMAAAAJ
Nanyang Technological University
Large Language ModelAI SafetyLLM Safety
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
248
 
H-index
8
 
i10-index
7
 
Publications
19
 
Co-authors
0
 
Contact
Emailzhouzhenhong@bupt.edu.cnTwitterOpen ↗GitHubOpen ↗
Publications
43 items
Feature-Aware Token Attack for Compression-Triggered Stealthy Failures in Large Vision-Language Models
2026
Cited
0
RESCUE: Repairing Language Model Errors to Sparse Circuits via Reinforcement Learning
2026
Cited
0
When Reasoning Goes Astray: Attention Dynamics of Uncontrolled Reasoning
2026
Cited
0
Uncovering Uncontrolled Repetition through Residual Stream Dynamics
2026
Cited
0
Long-Term Memory-Guided Enhancement for Target Perception in Audio-Language Models
2026
Cited
0
Still There, No Longer Seen: Exposing Compression-Induced Risk in Large Vision-Language Models
2026
Cited
0
In-Place Instruction Following in Diffusion Language Models
2026
Cited
0
From Inaudible Inputs to Model Failures: Low-Frequency Safety Risks in LALMs
2026
Cited
0
Resume
Academic Achievements
  • ICLR 2025 (Oral, Top 1.8%): 'On the Role of Attention Heads in Large Language Model Safety'
  • ICML 2025: Paper accepted
  • ACL 2025: Paper accepted
  • EMNLP 2024: Two papers accepted, one as co-first author
  • AAAI 2024: 'Quantifying and Analyzing Entity-Level Memorization in Large Language Models'
  • arXiv preprint: 'Speak Out of Turn: Safety Vulnerability of Large Language Models in Multi-turn Dialogue'
  • Oct 2023: First-class Postgraduate Academic Scholarship, Beijing University of Posts and Telecommunications
  • Oct 2022: Second-class Postgraduate Academic Scholarship, Beijing University of Posts and Telecommunications