Research · Chinese Neologism Safety
CMU Qatar · Nov 2023 – Jan 2024
Designed a 5k-entry dataset of harmful Chinese neologisms for LLaMA-7B fine-tuning. +20% F1 on adversarial text vs baseline LLMs.
This is the plain-HTML mirror served to crawlers, LLMs, and curl. Humans with a JavaScript-enabled browser see the rich React/XP-themed SPA at the same URL.