aiqbench analysis
BeliefGraph-JEPA advances time-series forecasting by integrating structured latent world models that explicitly track how future actions influence multiple targets through evolving relational graphs. This approach outperforms standard baselines in diverse domains like clinical and industrial systems by separating dynamic latent effects from direct driver information. The work represents a significant step toward more interpretable and robust predictive architectures for complex, action-conditioned environments.
BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series
arXiv:2610.05409v1 Announce Type: cross Abstract: Action-conditioned time-series forecasting requires accounting for how future actions and exogenous forcings influence multiple targets through partially observed effects with different delays and persistence. Direct conditioning leaves the evolution and target-specific influence of these effects implicit in the predictor, while static relational graphs specify connections without tracking evolving effects. This motivates representing future-driver influence through structured latent states that evolve over the forecast horizon and route information to individual targets. We introduce BeliefGraph-JEPA, a structured latent world model that factorizes driver influence into typed latent-effect states. These states are rolled forward under future drivers and routed through a graph to target-specific nodes, forming the predictive base of a joint-embedding predictive architecture. A capacity-controlled residual supplements this base with direct driver information. On four multi-target clinical, agricultural, environmental, and industrial systems, the framework outperforms a range of pretrained and supervised known-future-covariate baselines. Matched controls isolate latent dynamics, future rollout, graph routing, and residual capacity; future rollout and graph-first residual routing improve forecasting across all four systems.
Abstract:Action-conditioned time-series forecasting requires accounting for how future actions and exogenous forcings influence multiple targets through partially observed effects with different delays and persistence. Direct conditioning leaves the evolution and target-specific influence of these effects implicit in the predictor, while static relational graphs specify connections without tracking evolving effects. This motivates representing future-driver influence through structured latent states that evolve over the forecast horizon and route information to individual targets. We introduce BeliefGraph-JEPA, a structured latent world model that factorizes driver influence into typed latent-effect states. These states are rolled forward under future drivers and routed through a graph to target-specific nodes, forming the predictive base of a joint-embedding predictive architecture. A capacity-controlled residual supplements this base with direct driver information. On four multi-target clinical, agricultural, environmental, and industrial systems, the framework outperforms a range of pretrained and supervised known-future-covariate baselines. Matched controls isolate latent dynamics, future rollout, graph routing, and residual capacity; future rollout and graph-first residual routing improve forecasting across all four systems.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2610.05409 [cs.LG]
(or arXiv:2610.05409v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.05409 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 262 words
Related coverage
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · Sep 25, 2026
I Act Therefore I Am: When Is JEPA's Action-Conditioning Enough to Learn Causal Mechanisms?
arXiv · Sep 28, 2026
LeHami: Symplectic Latent Dynamics for World Models
arXiv · Oct 15, 2026
Latent Video Prediction for World Modeling: An Evaluation Uncovering Intriguing Favorable Evidence
arXiv · Oct 1, 2026