Adam Hazimeh
Titles as the sources gave them. In 1 entry, with 0 checked quotations.
In the Base
Every entry they appear in.
Beyond the Prompt: Leveraging Pre-Decoding States for Jailbreak Detection in dLLMs
According to the abstract, Hazimeh and colleagues propose ReFuse, an inference-time detector for jailbreaks in diffusion language models that combines classifiers on prompt representations and on hidden states of masked response positions before any tokens are generated.