Scholar
Kimin Lee
Google Scholar ID: 92M8xv4AAAAJ
KAIST
Citations & Impact
Citations
14,069
H-index
38
i10-index
56
Publications
20
Co-authors
158
list available
Publications
Cited
0
Resume (English only)
Academic Achievements
- Published at EMNLP 2025: 'Improbable Bigrams Expose Vulnerabilities of Incomplete Tokens in Byte-Level Tokenizers'
- Published at CoLLAs 2025: 'Benchmarking Mobile Device Control Agents across Diverse Configurations'
- Multiple papers at ACL 2025 main conference, including works on human feedback influence, long-context vulnerabilities, and LLM agent safety
- CVPR 2025 paper: 'Silent Branding Attack: Trigger-free Data Poisoning Attack on Text-to-Image Diffusion Models'
- USENIX Security 2025: 'When LLMs Go Online: The Emerging Threat of Web-Enabled LLMs'
- Multiple ICLR 2025 publications on LLM agents, diffusion model alignment, visual reward learning, and preference annotation (one accepted as oral presentation, top 1.77%)
- Active contributions in LLM alignment, safety evaluation, diffusion model security, and agent decision-making
