发表机构
LawZero; Mila – Quebec AI Institute; McGill University（LawZero; 米拉 - 魁北克人工智能研究所; 麦吉尔大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究使用朗之万动力学训练模型时避免轨迹进入不安全集的问题，通过分析强凸损失及相关失败区域，得出平衡质量、轨迹概率等界，引入局部松弛率，揭示强凸性与不安全集形状对训练轨迹的影响。
AI 中文摘要
使用有噪声梯度下降训练模型可理想化地视为损失景观上的过阻尼朗之万动力学，一个自然的安全问题是限制轨迹位于指定失败区域$\mathcal{A}_H$的概率$\nu_t(\mathcal{A}_H)=\mathbb{P}(Q_t\in\mathcal{A}_H)$。我们针对$d$维光滑、强凸损失以及与最小值由能量间隙隔开的失败区域进行研究。得出三个界。训练结束时，平衡质量$\pi(\mathcal{A}_H)$在$d$中呈指数小，噪声小时有互补能量障碍率。沿轨迹，一个无形状界$\nu_t(\mathcal{A}_H)\leq\pi(\mathcal{A}_H)(1+\sqrt{\chi_0^2/\pi(\mathcal{A}_H)}e^{-mt})$表明，经过约$d$阶的预热时间后，集内概率松弛到（两倍）静态值，仅使用损失的全局谱隙$m$。一个奥恩斯坦 - 乌伦贝克示例表明这种预热是必要的：平衡壳的角切片即使平衡质量很小也可能瞬间膨胀$d$的指数倍。为排除这种膨胀，我们引入与失败区域相关的局部松弛率，通过其中心指标的谱测度而非狄利克雷形式瑞利商来定义。对于几何上孤立的区域，该率超过全局率，按比例缩小预热时间，并与最大原理上限相结合，可及时统一限制轨迹概率。情况是强凸性决定训练松弛的速度，但不安全集的形状决定轨迹在返回时是否会穿过它。
英文摘要
Training a model with noisy gradient descent can be idealized as overdamped Langevin dynamics, and a natural safety question is to bound the probability $ν_t(\mathcal{A}_H) = \mathbb{P}(Q_t \in \mathcal{A}_H)$ that the trajectory lies in a designated failure region $\mathcal{A}_H$. We study this for a smooth, strongly convex loss in $d$ dimensions, with $\mathcal{A}_H$ separated from the minimizer by an energy gap. At the end of training, the equilibrium mass $π(\mathcal{A}_H)$ is exponentially small in $d$, with a complementary energy-barrier rate when the noise is small. Along the trajectory, a shape-free bound $ν_t(\mathcal{A}_H) \le π(\mathcal{A}_H)(1 + \sqrt{χ_0^2/π(\mathcal{A}_H)}\,e^{-mt})$ shows the in-set probability relaxes to (twice) the static value after a burn-in of order $d$, using only the global spectral gap $m$. A worked Ornstein-Uhlenbeck example shows this burn-in is necessary: an angular slice of the equilibrium shell can transiently swell by a factor exponential in $d$, though its equilibrium mass is tiny. To rule this out we introduce a local relaxation rate, defined through the spectral measure of the region's centered indicator rather than a Dirichlet-form Rayleigh quotient. For geometrically isolated regions this rate exceeds the global one, shrinking the burn-in, and with a maximum-principle ceiling it caps the trajectory probability uniformly in time. Strong convexity sets how fast training relaxes, but the shape of the unsafe set decides whether the trajectory bulges through it on the way to equilibrium.
Commentsv2: expanded and revised; 21 pages, 3 figures
