Abstract
Generative systems for dynamic interactive environments may face technical challenges related to maintaining logical consistency and narrative coherence, sometimes referred to as state drift. A computer-implemented system, such as a server or cloud computing platform, can utilize an orchestration layer to coordinate multiple specialized generative pipelines, for example, for narrative, visuals, and audio. A technique can involve constraining a generative language model to output a structured data object that separates a persistent world state vector from an ephemeral narrative payload. The persistent state vector can be reinjected into the model on subsequent turns to help maintain long-term coherence, while metadata in the payload can be used to condition the visual and audio pipelines for multi-modal consistency. This approach may mitigate state drift and support a more cohesive user experience.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Singhal, Parnika; Saini, Chirag; and Thakur, Nilanjana, "System for Orchestrating Multi-Modal Generative Pipelines for State-Consistent Interactive Environments", Technical Disclosure Commons, (August 14, 2026)
https://www.tdcommons.org/dpubs_series/11376