Fang Wu

ADWIN introduces an adaptive window framework for on-policy distillation (OPD) that efficiently manages the supervision horizon by training on short, teacher-anchored prefixes while using delayed full-rollout probes to maintain alignment, significantly reducing training cost while preserving accuracy.

The paper investigates multimodal jailbreak robustness across various reasoning paradigms and finds that explicit image-tool interaction significantly improves safety by guiding the model's internal representations toward safer directions.

The paper investigates multimodal jailbreak robustness across various reasoning paradigms and finds that explicit image-tool interaction significantly improves safety by shifting the model's internal representations toward a safety-relevant direction.