通过检测视觉语言动作模型的依赖关系，实时发现其决策失误。
GUARD: Grounding Uncertainty and Ablation-Based Risk Detection for Diffusion-Based VLAs
- 在不修改预训练模型的前提下，分析视觉语言特征缓存中关键条目的影响。
- 在五种未见任务上平均提升5.73%的异常检测准确率，优于现有方法。
- 适用于多种模型、任务和场景，适合部署在真实机器人系统中。
基于扩散的视觉-语言-动作（VLA）策略即使在视觉与语言证据弱关联时也能生成合理动作。本文提出GUARD，一种测试阶段故障检测方法，可在不修改预训练策略的情况下评估其对多模态输入的依赖程度。GUARD通过分析最终视觉-语言模型键值缓存中各条目的影响，构造消融后的反事实缓存，并对比其去噪响应与原始条件的差异，从而生成包含敏感度、注意力熵、模态偏差和接地效率的诊断流。这些指标在线校准后由轻量级时序分类器处理。在五个策略基准设置下，使用Pi0、SmolVLA和Alpamayo-1.5在LIBERO、SimplerEnv、MetaWorld和PhysicalAI-AV数据集上进行任务未见分割评估，GUARD在四个未见任务设置上取得最佳ROC-AUC，在剩余一项中排名第二，相比最强竞品运行时监控方法平均提升5.73个百分点，同时保持在最佳已见任务平均值0.19点以内。结果表明，直接探测动作头对多模态证据的依赖性可提供跨模型、任务、体态与领域的可迁移故障信号。
原文摘要 · Abstract (English)
Diffusion-based vision-language-action (VLA) policies can generate plausible actions even when their predictions are weakly grounded in the visual and language evidence defining the task. We introduce GUARD, a test-time failure detection method that measures this grounding without modifying the pretrained policy. GUARD estimates the influence of token-indexed entries in the final vision-language model key-value (KV) cache, constructs counterfactual caches by ablating salient KV entries, and compares their denoising responses with the original conditioning. Based on the comparison, we derive GUARD diagnostic stream including sensitivity, attention entropy, modality bias, and grounding efficiency, which are calibrated online and processed by a lightweight temporal classifier. We evaluate GUARD under task-held-out splits across five policy-benchmark settings, using Pi0, SmolVLA, and Alpamayo-1.5 on LIBERO, SimplerEnv, MetaWorld, and PhysicalAI-AV. GUARD achieves the best ROC-AUC on four of five unseen-task settings and ranks second on the remaining setting, improving the average unseen-task ROC-AUC by 5.73 percentage points over the strongest competing runtime monitor while remaining within 0.19 points of the best seen-task average. These results show that directly probing action-head dependence on multimodal evidence provides a transferable failure signal across policies, tasks, embodiments, and domains.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。