Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces a parameter-efficient continuous-time generative modeling framework based on controlled vector fields, where a small set of fixed vector fields are modulated by learned scalar control functions to construct expressive flows. By leveraging bracket-generating fixed fields, the method achieves expressive transport with fewer learned parameters compared to standard vector-field parameterizations. The authors demonstrate the framework's ability to transport distributions via a continuous-normalizing-flow likelihood objective, validated through experiments on synthetic distributions.
Unlock expressive generative modeling with far fewer parameters by steering fixed vector fields with learned scalar controls, offering a geometric shortcut to high-dimensional flows.
We introduce a continuous-time generative modeling framework, motivated by the Chow-Rashevskii theorem, that builds expressive flows from a small set of fixed vector fields and learned scalar controls. Instead of learning an unconstrained high-dimensional vector field, our framework constructs the velocity by modulating fixed vector fields with learned scalar control functions. When the fixed fields are bracket-generating, their Lie algebra spans the ambient space, providing a mechanism for expressive transport with only a small number of learned control channels and offering a parameter-efficient geometric alternative to standard vector-field parameterizations. This decoupled formulation yields a structured and interpretable generative model in which the number of learned scalar output channels can be chosen independently of the ambient dimension. We formulate an expressivity principle showing that, under suitable controllability and well-posedness assumptions, such controlled flows can transport a source distribution to a target distribution. We train the resulting model using a continuous-normalizing-flow likelihood objective and present proof-of-concept experiments on synthetic distributions.