Weikai Lin
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Research Timeline
2026
SafeDream: Safety World Model for Proactive Early Jailbreak Detection
SAFEDREAM introduces a lightweight, external world-model framework that proactively detects multi-turn jailbreak attacks by modeling cumulative safety erosion and predicting early failure points.
Highlighted terms show continued research focus across papers