原文：Causal world models under physical parameter shifts
作者：Chu Lin, Yi Sun
来源：arXiv cs.RO（机器人）
正文
Computer Science > Machine Learning
arXiv:2610.29088v1 (cs)
[Submitted on 21 Oct 2026]
Title:Causal world models under physical parameter shifts
Authors:Chu Lin, Yi Sun
View a PDF of the paper titled Causal world models under physical parameter shifts, by Chu Lin and Yi Sun
View PDF
HTML (experimental)
Abstract:World models predict how embodied systems evolve under actions, but their accuracy can degrade when physical conditions move outside the training distribution. We study two components of causal modeling: which inputs a predictor may use, and how its mechanism computes their effects. In a synthetic benchmark with shortcut features, a frozen dependency mask derived from paired interventions reduces mean out-of-distribution (OOD) RMSE from 1.769 for a fully connected predictor to 0.697, close to the generator-oracle mask (0.694). For robot link-mass extrapolation in CausalWorld, with disjoint training and test mass ranges, we develop finger-local world models in which link masses enter inertia and dynamics-bias terms through explicit affine coefficients, actions act through either a learned torque network or the known PD controller, and learned residuals capture contact-related effects. Compared with a separately trained control whose mass inputs are fixed at training means, the known-PD model turns out-of-range masses into a consistent predictive gain: median time-averaged joint R2 over 150-step rollouts rises from 0.609 to 0.727, with improvement in all ten task-seed runs (exact two-sided sign test, p = 0.002). In a shared 100-step forecast, the learned-torque model is competitive with the strongest recurrent robotic world model (RWM) configuration (median joint R2 of 0.755 - 0.768 versus 0.752), whereas supplying out-of-range masses sharply degrades the MLP and recurrent predictors. These results indicate that how a physical parameter enters the dynamics mechanism determines whether its out-of-range value helps or harms prediction.
Comments:
2 figures
Subjects:
Machine Learning (cs.LG); Robotics (cs.RO)
Cite as:
arXiv:2610.29088 [cs.LG]
(or
arXiv:2610.29088v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.29088
Focus to learn more
arXiv-issued DOI via DataCite (pending registration)
主题
机器学习 · 机器人
由「前沿雷达」于 2026-10-21 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。
评论
加载中…