Saved in:
Bibliographic Details
Main Authors: Le, Thien, Weber, Melanie
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.20074
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917511536050176
author Le, Thien
Weber, Melanie
author_facet Le, Thien
Weber, Melanie
contents Distillation transfers knowledge from a large model trained on broad data to a smaller, more efficient model suitable for deployment. In structured prediction settings, prior knowledge about the task can guide the choice of a target architecture that is algorithmically aligned with the underlying problem. Building on recent learning-theoretic analyses of decision-tree (DT) distillation (Boix-Adsera, 2024), we study when distillation succeeds for combinatorial optimization tasks. We focus on the case where the target model is a graph neural network whose architecture is aligned with a dynamic programming (DP) algorithm for the task. Assuming that the source model is sufficiently rich, formalized through the linear representation hypothesis (LRH) (Elhage et al., 2022; Park et al., 2024), we show that the distillation problem can be solved efficiently in the complexity parameters of the DP transition function, represented as a DT. Our results provide a rigorous sufficient condition for successful distillation in the flavour of algorithmic alignment.
format Preprint
id arxiv_https___arxiv_org_abs_2605_20074
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Towards Distillation Guarantees under Algorithmic Alignment for Combinatorial Optimization
Le, Thien
Weber, Melanie
Machine Learning
Distillation transfers knowledge from a large model trained on broad data to a smaller, more efficient model suitable for deployment. In structured prediction settings, prior knowledge about the task can guide the choice of a target architecture that is algorithmically aligned with the underlying problem. Building on recent learning-theoretic analyses of decision-tree (DT) distillation (Boix-Adsera, 2024), we study when distillation succeeds for combinatorial optimization tasks. We focus on the case where the target model is a graph neural network whose architecture is aligned with a dynamic programming (DP) algorithm for the task. Assuming that the source model is sufficiently rich, formalized through the linear representation hypothesis (LRH) (Elhage et al., 2022; Park et al., 2024), we show that the distillation problem can be solved efficiently in the complexity parameters of the DP transition function, represented as a DT. Our results provide a rigorous sufficient condition for successful distillation in the flavour of algorithmic alignment.
title Towards Distillation Guarantees under Algorithmic Alignment for Combinatorial Optimization
topic Machine Learning
url https://arxiv.org/abs/2605.20074