A new differentiable layer with convex parameter space universally approximates generalized convex functions and their gradients, enabling single-level reformulations of bilevel problems in optimal transport and multi-good auctions.
Backpropaga- tion applied to handwritten zip code recognition,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Universal Representation of Generalized Convex Functions and their Gradients
A new differentiable layer with convex parameter space universally approximates generalized convex functions and their gradients, enabling single-level reformulations of bilevel problems in optimal transport and multi-good auctions.