PolyILR supplies a canonical orthonormal decomposition of the Aitchison tangent space aligned with arbitrary tree structures.
Simplex-to-euclidean bijections for categorical flow matching
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
A geometric latent-subspace model on Riemannian manifolds of categorical distributions enables low-dimensional generative modeling of discrete data via isometries and geometric PCA for flow matching.
Continuous flows on token embeddings with flow-map distillation produce one-step language models whose quality exceeds recent 8-step discrete diffusion baselines on LM1B and OpenWebText.
citing papers explorer
-
Tree-Structured Orthonormal Decomposition of the Aitchison Simplex
PolyILR supplies a canonical orthonormal decomposition of the Aitchison tangent space aligned with arbitrary tree structures.
-
Generative Modeling of Discrete Data Using Geometric Latent Subspaces
A geometric latent-subspace model on Riemannian manifolds of categorical distributions enables low-dimensional generative modeling of discrete data via isometries and geometric PCA for flow matching.
-
Flow Map Language Models: One-step Language Modeling via Continuous Denoising
Continuous flows on token embeddings with flow-map distillation produce one-step language models whose quality exceeds recent 8-step discrete diffusion baselines on LM1B and OpenWebText.