Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Yankai, Zhang, Hanrong, He, Bowei, Yu, Philip S., Xue, Liu
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2605.30089
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914614374039552
author Chen, Yankai
Zhang, Hanrong
He, Bowei
Yu, Philip S.
Xue
Liu
author_facet Chen, Yankai
Zhang, Hanrong
He, Bowei
Yu, Philip S.
Xue
Liu
contents Standard Set Representation Learning methods typically excel on curated data but often overlook the challenge of inference-time element corruption. This refers to scenarios where deployed models encounter element-level degradations, such as outliers or missing components, that may distort set representation and degrade performance. We propose SW-DRSO, a distributionally robust optimization framework tailored for sets. Rather than minimizing loss solely on observed training data, SW-DRSO optimizes a tractable surrogate of the worst-case expected loss over a family of plausible inference-time variations. We introduce a barycentric adversary that approximates the intractable search over corrupted sets by a differentiable training-time optimization over simplex weights. Extensive experiments across four tasks demonstrate that SW-DRSO effectively enhances robustness against corruption while maintaining high overall performance.
format Preprint
id arxiv_https___arxiv_org_abs_2605_30089
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Distributionally Robust Set Representation Learning Under Inference-Time Element Corruption
Chen, Yankai
Zhang, Hanrong
He, Bowei
Yu, Philip S.
Xue
Liu
Machine Learning
Standard Set Representation Learning methods typically excel on curated data but often overlook the challenge of inference-time element corruption. This refers to scenarios where deployed models encounter element-level degradations, such as outliers or missing components, that may distort set representation and degrade performance. We propose SW-DRSO, a distributionally robust optimization framework tailored for sets. Rather than minimizing loss solely on observed training data, SW-DRSO optimizes a tractable surrogate of the worst-case expected loss over a family of plausible inference-time variations. We introduce a barycentric adversary that approximates the intractable search over corrupted sets by a differentiable training-time optimization over simplex weights. Extensive experiments across four tasks demonstrate that SW-DRSO effectively enhances robustness against corruption while maintaining high overall performance.
title Distributionally Robust Set Representation Learning Under Inference-Time Element Corruption
topic Machine Learning
url https://arxiv.org/abs/2605.30089