Diverging Flows: Detecting Extrapolations in Conditional Generation

TL;DR

Diverging Flows enhances flow models by enabling simultaneous conditional generation and reliable detection of off-manifold extrapolations, improving safety and trustworthiness in critical applications.

Contribution

It introduces a novel method that structurally enforces inefficient transport for off-manifold inputs, allowing detection of extrapolations without sacrificing predictive accuracy.

Findings

Effective detection of off-manifold inputs across various tasks

Maintains predictive fidelity and inference speed

Applicable to domains like medicine, robotics, and climate science

Abstract

The ability of Flow Matching (FM) to model complex conditional distributions has established it as the state-of-the-art for prediction tasks (e.g., robotics, weather forecasting). However, deployment in safety-critical settings is hindered by a critical extrapolation hazard: driven by smoothness biases, flow models yield plausible outputs even for off-manifold conditions, resulting in silent failures indistinguishable from valid predictions. In this work, we introduce Diverging Flows, a novel approach that enables a single model to simultaneously perform conditional generation and native extrapolation detection by structurally enforcing inefficient transport for off-manifold inputs. We evaluate our method on synthetic manifolds, cross-domain style transfer, and weather temperature forecasting, demonstrating that it achieves effective detection of extrapolations without compromising predictive fidelity or inference latency. These results establish Diverging Flows as a robust solution for trustworthy flow models, paving the way for reliable deployment in domains such as medicine, robotics, and climate science.