PolicyShiftGuard: Benchmarking and Improving Policy-Adaptive Image Guardrails

  • 类型:arxiv
  • 标识:2607.05910
  • 链接:https://arxiv.org/abs/2607.05910
  • 主分类:risk
  • 形态:benchmark
  • 被引:0
  • 被引来源:Semantic Scholar + OpenAlex
  • S2被引:0
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:This work proposes PolicyShiftGuard, a compact policy-conditioned guardrail trained with a two-stage training recipe that combines Randomized Policy SFT (RP-SFT) with Boundary-Pair Policy Adaptation (BP-Adapt), and confirms that matched pass/block boundary pairs are essential for stable policy adaptation.
  • OpenAlex ID:W7167679369
  • OpenAlex DOI:10.48550/arxiv.2607.05910
  • DOI:10.48550/arxiv.2607.05910
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2607.05910
  • OpenAlex更新:2026-07-19
  • 副分类:evaluation
  • 待LLM分类:否
  • 标题中文:PolicyShiftGuard: Benchmarking and Improving Policy-Adaptive Image Guardrails
  • TLDR中文:论文提出 PolicyShiftGuard,一个紧凑的策略条件护栏,采用结合随机策略 SFT(RP-SFT)与边界对策略适配(BP-Adapt)的两阶段训练方案,并验证匹配的通过/拒绝边界对是稳定策略适配的关键。
  • 来源文件
  • /inbox/tom/_candidates/2026-07-16-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]