AI safety researcher

Hongwei Li

University of Electronic Science and Technology of China

Google Scholar dblp ORCID

Last 12 months12 papers

Defenses 1Attacks 1

AI safety papers from the last year

On this site: 34 papers

September 2026

FragToken: Amplifying LLM Inference Costs through Noncanonical Token Generation
Sep 25arXiv ↗
RAPID: A Real-Time Defense Against Unauthorized Model Distillation for Text-to-Image Services
Sep 14DefensesarXiv ↗
CounterPersona: Append-Only Defense Against Unauthorized Persona Skill Distillation
Sep 14arXiv ↗

August 2026

July 2026

InkShield: Writing Style Protection Against Unauthorized Handwriting Mimicry
Jul 29arXiv ↗

April 2026

February 2026

BadTemplate: A Training-Free Backdoor Attack via Chat Template Against Large Language Models
Feb 5AttacksarXiv ↗

January 2026

State Backdoor: Towards Stealthy Real-world Poisoning Attack on Vision-Language-Action Model in State Space
Jan 7arXiv ↗

November 2025

TEAR: Temporal-aware Automated Red-teaming for Text-to-Video Models
Nov 26arXiv ↗

Authorship follows dblp and the author's own Google Scholar profile; only AI safety papers on this site are listed.