4 papers
Branching Flows: Discrete, Continuous, and Manifold Flow Matching with Splits and Deletions
Lukas Billera, Hedwig Nora Nordlinder, Jack Collier Ryder +4
Diffusion and flow matching approaches to generative modeling have shown promise in domains where the state space is continuous, such as image generation or protein folding & desig…
Latent Process Generator Matching
Lukas Billera, Hedwig Nora Nordlinder, Ben Murrell
Many recent flow-matching and diffusion-style generative models rely on auxiliary stochastic dynamics during training: a richer process is simulated to define conditional targets,…
Generative vector search to improve pathology foundation models across multimodal vision-language tasks
Markus Ekvall, Ludvig Bergenstråhle, Patrick Truong +2
Retrieval-augmented generation improves large language models by grounding outputs in external knowledge sources, reducing hallucinations and addressing knowledge cutoffs. However,…
Time dependent loss reweighting for flow matching and diffusion models is theoretically justified
Lukas Billera, Hedwig Nora Nordlinder, Ben Murrell
This brief note clarifies that, in Generator Matching (which subsumes a large family of flow matching and diffusion models over continuous, manifold, and discrete spaces), both the…