Model Agreement via Anchoring

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Eaton, Eric, Goel, Surbhi, Hussing, Marcel, Kearns, Michael, Roth, Aaron, Sengupta, Sikata Bela, Sorrell, Jessica
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908854848061440
author Eaton, Eric
Goel, Surbhi
Hussing, Marcel
Kearns, Michael
Roth, Aaron
Sengupta, Sikata Bela
Sorrell, Jessica
author_facet Eaton, Eric
Goel, Surbhi
Hussing, Marcel
Kearns, Michael
Roth, Aaron
Sengupta, Sikata Bela
Sorrell, Jessica
contents Numerous lines of aim to control $\textit{model disagreement}$ -- the extent to which two machine learning models disagree in their predictions. We adopt a simple and standard notion of model disagreement in real-valued prediction problems, namely the expected squared difference in predictions between two models trained on independent samples, without any coordination of the training processes. We would like to be able to drive disagreement to zero with some natural parameter(s) of the training procedure using analyses that can be applied to existing training methodologies. We develop a simple general technique for proving bounds on independent model disagreement based on $\textit{anchoring}$ to the average of two models within the analysis. We then apply this technique to prove disagreement bounds for four commonly used machine learning algorithms: (1) stacked aggregation over an arbitrary model class (where disagreement is driven to 0 with the number of models $k$ being stacked) (2) gradient boosting (where disagreement is driven to 0 with the number of iterations $k$) (3) neural network training with architecture search (where disagreement is driven to 0 with the size $n$ of the architecture being optimized over) and (4) regression tree training over all regression trees of fixed depth (where disagreement is driven to 0 with the depth $d$ of the tree architecture). For clarity, we work out our initial bounds in the setting of one-dimensional regression with squared error loss -- but then show that all of our results generalize to multi-dimensional regression with any strongly convex loss.
format Preprint
id arxiv_https___arxiv_org_abs_2602_23360
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Model Agreement via Anchoring
Eaton, Eric
Goel, Surbhi
Hussing, Marcel
Kearns, Michael
Roth, Aaron
Sengupta, Sikata Bela
Sorrell, Jessica
Machine Learning
Artificial Intelligence
Numerous lines of aim to control $\textit{model disagreement}$ -- the extent to which two machine learning models disagree in their predictions. We adopt a simple and standard notion of model disagreement in real-valued prediction problems, namely the expected squared difference in predictions between two models trained on independent samples, without any coordination of the training processes. We would like to be able to drive disagreement to zero with some natural parameter(s) of the training procedure using analyses that can be applied to existing training methodologies. We develop a simple general technique for proving bounds on independent model disagreement based on $\textit{anchoring}$ to the average of two models within the analysis. We then apply this technique to prove disagreement bounds for four commonly used machine learning algorithms: (1) stacked aggregation over an arbitrary model class (where disagreement is driven to 0 with the number of models $k$ being stacked) (2) gradient boosting (where disagreement is driven to 0 with the number of iterations $k$) (3) neural network training with architecture search (where disagreement is driven to 0 with the size $n$ of the architecture being optimized over) and (4) regression tree training over all regression trees of fixed depth (where disagreement is driven to 0 with the depth $d$ of the tree architecture). For clarity, we work out our initial bounds in the setting of one-dimensional regression with squared error loss -- but then show that all of our results generalize to multi-dimensional regression with any strongly convex loss.
title Model Agreement via Anchoring
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.23360