FlashDiffusion: Fused Tiled Kernel Spectral Decomposition
cs.LG
Submitted: 2026-09-18
Updated: 2026-09-18
Sources
- An Improved Approximation Algorithm for the Column Subset Selection Problem
- Diffusion Map Autoencoder
- Linearized Diffusion Map
- The Diffusion-Attention Connection
- Kernel Operations on the GPU, with Autodiff, without Memory Overflows
- Spectrum of the Laplace-Beltrami Operator and the Phase Structure of Causal Dynamical Triangulation
- FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
- Graph Laplacians and their convergence on random neighborhood graphs
- Fast Kernel Summation in High Dimensions via Slicing and Fourier Transforms
- Landmark Diffusion Maps (L-dMaps): Accelerated manifold learning out-of-sample extension
- Efficient and robust approximate nearest neighbor search using Hierarchical Navigable Small World graphs
- megaman: Manifold Learning with Millions of points
- Online normalizer calculation for softmax
- Markov State Models from short non-Equilibrium Simulations - Analysis and Correction of Estimation Bias
- An Analysis of the Convergence of Graph Laplacians
- Time-lagged autoencoders: Deep learning of slow collective variables for molecular kinetics
- FlashAttention-4: Algorithm and Kernel Pipelining Co-Design for Asymmetric Hardware Scaling
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks