[2603.14354] Deconfounded Lifelong Learning for Autonomous Driving via Dynamic Knowledge Spaces

Computer Science > Machine Learning arXiv:2603.14354 (cs) [Submitted on 15 Mar 2026 (v1), last revised 30 Mar 2026 (this version, v2)] Title:Deconfounded Lifelong Learning for Autonomous Driving via Dynamic Knowledge Spaces Authors:Jiayuan Du, Yuebing Song, Yiming Zhao, Xianghui Pan, Jiawei Lian, Yuchu Lu, Liuyi Wang, Chengju Liu, Qijun Chen View a PDF of the paper titled Deconfounded Lifelong Learning for Autonomous Driving via Dynamic Knowledge Spaces, by Jiayuan Du and 8 other authors View PDF HTML (experimental) Abstract:End-to-End autonomous driving (E2E-AD) systems face challenges in lifelong learning, including catastrophic forgetting, difficulty in knowledge transfer across diverse scenarios, and spurious correlations between unobservable confounders and true driving intents. To address these issues, we propose DeLL, a Deconfounded Lifelong Learning framework that integrates a Dirichlet process mixture model (DPMM) with the front-door adjustment mechanism from causal inference. The DPMM is employed to construct two dynamic knowledge spaces: a trajectory knowledge space for clustering explicit driving behaviors and an implicit feature knowledge space for discovering latent driving abilities. Leveraging the non-parametric Bayesian nature of DPMM, our framework enables adaptive expansion and incremental updating of knowledge without predefining the number of clusters, thereby mitigating catastrophic forgetting. Meanwhile, the front-door adjustment mechanism utilizes t...