A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Paper proves autodiff systems are correct for non-differentiable functions.
Proposes efficient, modular method for implicit differentiation.
AutoDiff combines auto-encoder and diffusion model for realistic tabular data synthesis.
Simplifies convolutions using tensor networks and einsum for efficient second-order methods.
New method reconstructs missing variables in time series using autoencoders and automatic differentiation.
Derivatives, mostly in the form of gradients and Hessians, are ubiquitous in machine learning. Automatic differentiation (AD), also called algorithmic differentiation or simply "autodiff", is a family of techniques similar to but more general than backpropagation for efficiently and accurately evaluating derivatives of…
Extends gradient-based optimization to spline functions.
Deep GMRFs improve spatial data modeling and prediction.
A new method uses natural gradients for efficient distribution optimization.
Study on numerical reliability of AD for MaxPool in neural nets.