Saved in:
Bibliographic Details
Main Authors: Modarres, Mohammad Reza, Abbasi, Sina, Pilehvar, Mohammad Taher
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.09642
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910648152096768
author Modarres, Mohammad Reza
Abbasi, Sina
Pilehvar, Mohammad Taher
author_facet Modarres, Mohammad Reza
Abbasi, Sina
Pilehvar, Mohammad Taher
contents Advances in dataset analysis techniques have enabled more sophisticated approaches to analyzing and characterizing training data instances, often categorizing data based on attributes such as ``difficulty''. In this work, we introduce RepMatch, a novel method that characterizes data through the lens of similarity. RepMatch quantifies the similarity between subsets of training instances by comparing the knowledge encoded in models trained on them, overcoming the limitations of existing analysis methods that focus solely on individual instances and are restricted to within-dataset analysis. Our framework allows for a broader evaluation, enabling similarity comparisons across arbitrary subsets of instances, supporting both dataset-to-dataset and instance-to-dataset analyses. We validate the effectiveness of RepMatch across multiple NLP tasks, datasets, and models. Through extensive experimentation, we demonstrate that RepMatch can effectively compare datasets, identify more representative subsets of a dataset (that lead to better performance than randomly selected subsets of equivalent size), and uncover heuristics underlying the construction of some challenge datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2410_09642
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle RepMatch: Quantifying Cross-Instance Similarities in Representation Space
Modarres, Mohammad Reza
Abbasi, Sina
Pilehvar, Mohammad Taher
Computation and Language
Advances in dataset analysis techniques have enabled more sophisticated approaches to analyzing and characterizing training data instances, often categorizing data based on attributes such as ``difficulty''. In this work, we introduce RepMatch, a novel method that characterizes data through the lens of similarity. RepMatch quantifies the similarity between subsets of training instances by comparing the knowledge encoded in models trained on them, overcoming the limitations of existing analysis methods that focus solely on individual instances and are restricted to within-dataset analysis. Our framework allows for a broader evaluation, enabling similarity comparisons across arbitrary subsets of instances, supporting both dataset-to-dataset and instance-to-dataset analyses. We validate the effectiveness of RepMatch across multiple NLP tasks, datasets, and models. Through extensive experimentation, we demonstrate that RepMatch can effectively compare datasets, identify more representative subsets of a dataset (that lead to better performance than randomly selected subsets of equivalent size), and uncover heuristics underlying the construction of some challenge datasets.
title RepMatch: Quantifying Cross-Instance Similarities in Representation Space
topic Computation and Language
url https://arxiv.org/abs/2410.09642