FlowBender introduces closed-loop training that lets conditional flow models learn correction policies from their own task-specific alignment errors, outperforming supervised and guidance baselines on fidelity and plausibility.
Inv- fusion: Bridging supervised and zero-shot diffusion for inverse problems.arXiv preprint arXiv:2504.01689, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FlowBender: Feedback-Aware Training for Self-Correcting Conditional Flows
FlowBender introduces closed-loop training that lets conditional flow models learn correction policies from their own task-specific alignment errors, outperforming supervised and guidance baselines on fidelity and plausibility.