等变世界模型的共形轨道有效信任视界
Conformal Orbit-Valid Trust Horizons for Equivariant World Models
AI总结:
针对具有已知群对称性的潜在世界模型,提出一种共形校准的信任视界认证方法,利用等变性实现轨道常数认证,实验验证了其保守性和非空性。
摘要:
学习到的世界模型仅在其展开误差可控的视界内有用。我们研究了具有已知群对称性的潜在世界模型的信任视界认证。给定一步潜在残差和有限时间展开估计,我们形成原始视界曲线,并用分裂共形乘法因子进行校准。在可复现的审计集上,共形因子为 $\gamma_\alpha=1.0$:原始证书在审计协议下已经是保守的。在50次稳定审计中,我们观察到零反保守违规,对应违规率的精确二项式95%上限为5.8%。我们的主要结构结果是,精确等变性将校准后的信任视界曲线沿群轨道传输:当环境动力学、编码器、预测器、动作变换和潜在度量满足所述等变性/不变性条件时,展开误差和信任视界是轨道常数。实验上,实现的模型表现出较小的轨道传输残差,在14次轨道审计中中位数为1.1%,最大为4.1%。该证书也是非空的(中位认证到测量视界比为0.67)。证书级校准成本研究显示了两个互补区域。在对称的2D基底上,等变、普通和增强模型均从单个校准扇区获得轨道有效性——无分离,因为基底已使非等变基线近似轨道鲁棒。3D偏航审计显示了另一个区域:等变模型获得单扇区安全且非空的轨道有效证书,而健康的非等变基线则付出违规、松弛、尖锐或额外扇区成本。该证书是保守的分布性审计而非全局可达性保证,且证书引导的子目标间距在当前3D CEM-MPC行为层中未得到确认。
英文摘要:
Learned world models are useful only over horizons on which their rollout error remains controlled. We study trust-horizon certification for latent world models with known group symmetries. Given a one-step latent residual and a finite-time expansion estimate, we form a raw horizon curve and calibrate it with a split-conformal multiplicative factor. On the reproducible audit set, the conformal factor is $γ_α=1.0$: the raw certificate is already conservative under the audit protocol. Across 50 stable audits, we observe zero anti-conservative violations, corresponding to an exact-binomial 95% upper bound of 5.8% on the violation rate. Our main structural result is that exact equivariance transports a calibrated trust-horizon curve over the group orbit: when the environment dynamics, encoder, predictor, action transform, and latent metric satisfy the stated equivariance/invariance conditions, rollout errors and trust horizons are orbit-constant. Empirically, the implemented models exhibit small orbit-transport residuals, with median 1.1% and maximum 4.1% over 14 orbit audits. The certificate is also non-vacuous (median certified-to-measured horizon ratio 0.67). A certificate-level calibration-cost study shows two complementary regimes. On a symmetric 2D substrate, equivariant, plain, and augmented models are all orbit-valid from a single calibration sector -- no separation, because the substrate already makes non-equivariant baselines approximately orbit-robust. A 3D yaw audit shows the other regime: the equivariant model obtains a one-sector safe and non-vacuous orbit-valid certificate, while healthy non-equivariant baselines pay violation, slack, sharpness, or additional-sector cost. The certificate is a conservative, distributional audit rather than a global reachability guarantee, and certificate-guided subgoal spacing is not confirmed in the current 3D CEM-MPC behavior layer.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。