研究驾驶大模型在传感器干扰下的推理稳定性，发现解释变化预示轨迹偏差激增。
Lost in Fog: Sensor Perturbations Expose Reasoning Fragility in Driving VLAs
- 通过8类传感器扰动测试10亿参数大模型，评估其推理一致性与轨迹可靠性关系。
- 解释变化时轨迹偏差达21.8米，是稳定时的5.3倍，相关性高达0.99。
- 生成因果链解释可提升轨迹准确率11.8%，适合关注自动驾驶安全性的研究者。
可解释自动驾驶规划依赖于生成的解释不仅有效，更需在真实传感器退化下保持可靠。本文对视觉-语言-动作（VLA）模型Alpamayo R1（10B参数）在1,996个场景下进行受控扰动实验，涵盖八类传感器扰动（四种高斯噪声强度、两种光照极端、两种雾度），共约18,000次推理测试。结果表明，推理一致性是轨迹可靠性的高保真指标：当因果链（CoC）解释在扰动后改变，轨迹偏差飙升至21.8米（原为4.1米），跨攻击类型相关系数r=0.99，样本级相关系数rpb=0.53（Cohen's d=1.12）。控制消融实验证明，在相同推理条件下，启用CoC生成可平均提升轨迹准确率11.8%（p < 0.0001）。在噪声强度σ∈{10,30,50,70}范围内，性能退化近似线性（R²=0.957），而常规预处理防御仅提供微弱缓解。这些结果确立了CoC一致性作为规划安全的量化代理，并推动基于推理的运行时监控以实现更安全的VLA部署。
原文摘要 · Abstract (English)
Interpretable autonomous driving planners depend not only on generating explanations, but also on those explanations remaining reliable under real-world sensor degradation. In this paper we present a controlled perturbation study of Vision-Language-Action (VLA) robustness in autonomous driving, evaluating Alpamayo R1 (10B parameters) across 1,996 scenarios under eight sensor perturbations (Gaussian noise at four intensities, two lighting extremes, and two fog levels; ${\sim}18{,}000$ inference trials). We find that reasoning consistency is a high-fidelity indicator of trajectory reliability: when Chain-of-Causation (CoC) explanations change after perturbation, trajectory deviation spikes $5.3{\times}$ (21.8m vs 4.1m), with $r\!=\!0.99$ across attack types and $r_{pb}\!=\!0.53$ per-sample (Cohen's $d\!=\!1.12$). A controlled ablation provides evidence that enabling CoC generation is associated with improved trajectory accuracy (11.8% on average across conditions; $p < 0.0001$) under matched inference settings. Over the tested noise range ($σ\in \{10, 30, 50, 70\}$), degradation is approximately linear ($R^2\!=\!0.957$), while standard input preprocessing defenses provide only marginal relief. Together, these results establish CoC consistency as a quantitative proxy for planning safety and motivate reasoning-based runtime monitoring for safer VLA deployment.
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。