Saved in:
Bibliographic Details
Main Authors: Amoh, Benjamin, Parker, Geoffrey G., Marrero, Wesley
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.12693
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918498383429632
author Amoh, Benjamin
Parker, Geoffrey G.
Marrero, Wesley
author_facet Amoh, Benjamin
Parker, Geoffrey G.
Marrero, Wesley
contents Decision-focused learning trains predictive models end-to-end against downstream decision loss, but online settings suffer delayed feedback: outcomes may not arrive for many environment interactions. We identify \emph{staleness amplification}, a failure mode unique to bilevel optimization under delay, in which gradient staleness couples with inner-solver sensitivity to inflate regret beyond single-level delay theory. We prove that any black-box delayed optimizer incurs an irreducible regret cost from inner-solver approximation error, and that gradient staleness contributes a quadratically growing transport error without bilevel-aware correction. Our algorithm, \textbf{IGT-OMD}, applies Implicit Gradient Transport to hypergradients within Online Mirror Descent, re-evaluating stale gradients at the current parameters using stored inner solutions. This method reduces transport error from a quadratic to a linear dependence on delay and achieves the first sublinear regret bound for delayed bilevel optimization with queue-length-adaptive step sizes. Controlled experiments provide a \emph{mechanistic fingerprint}: transport benefit is exactly $0.0\%$ ($p=1.00$) at unit delay and grows monotonically to $9.5\%$ at fifty rounds ($p<0.001$), isolating the correction's effect. On Linear Quadratic Regulator, Warcraft shortest-path, and Sinkhorn optimal transport, IGT-OMD reduces decision loss by $17$--$55\%$ relative to single-level baselines, with phase transitions matching the theory.
format Preprint
id arxiv_https___arxiv_org_abs_2605_12693
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle IGT-OMD: Implicit Gradient Transport for Decision-Focused Learning under Delayed Feedback
Amoh, Benjamin
Parker, Geoffrey G.
Marrero, Wesley
Machine Learning
68Q25, 90C25, 90C47
I.2.6; G.1.6
Decision-focused learning trains predictive models end-to-end against downstream decision loss, but online settings suffer delayed feedback: outcomes may not arrive for many environment interactions. We identify \emph{staleness amplification}, a failure mode unique to bilevel optimization under delay, in which gradient staleness couples with inner-solver sensitivity to inflate regret beyond single-level delay theory. We prove that any black-box delayed optimizer incurs an irreducible regret cost from inner-solver approximation error, and that gradient staleness contributes a quadratically growing transport error without bilevel-aware correction. Our algorithm, \textbf{IGT-OMD}, applies Implicit Gradient Transport to hypergradients within Online Mirror Descent, re-evaluating stale gradients at the current parameters using stored inner solutions. This method reduces transport error from a quadratic to a linear dependence on delay and achieves the first sublinear regret bound for delayed bilevel optimization with queue-length-adaptive step sizes. Controlled experiments provide a \emph{mechanistic fingerprint}: transport benefit is exactly $0.0\%$ ($p=1.00$) at unit delay and grows monotonically to $9.5\%$ at fifty rounds ($p<0.001$), isolating the correction's effect. On Linear Quadratic Regulator, Warcraft shortest-path, and Sinkhorn optimal transport, IGT-OMD reduces decision loss by $17$--$55\%$ relative to single-level baselines, with phase transitions matching the theory.
title IGT-OMD: Implicit Gradient Transport for Decision-Focused Learning under Delayed Feedback
topic Machine Learning
68Q25, 90C25, 90C47
I.2.6; G.1.6
url https://arxiv.org/abs/2605.12693