Papers
arxiv:2611.19286

On Representational and Functional Alignment of Training Data Attribution Methods for Language Models

Published on Nov 14
Authors:
,

Abstract

A framework directly compares training data attribution methods along functional alignment, representational alignment, and cross-method predictability, revealing low task-dependent cross-paradigm alignment and suggesting similar downstream performance stems from agreement on relative document importance rather than full training-data rankings.

Existing approaches to training data attribution derive influence estimates from different sources of information, including gradients, model-internal states, and surface-level similarity. However, comparisons between attribution methods remain incomplete, often relying on indirect evaluations through downstream task performance. In this work, we introduce a framework for directly comparing methods along three dimensions: functional alignment, which measures correlation between full attribution vectors; representational alignment, which captures structural similarity of the representations underlying attribution; and cross-method predictability, which evaluates agreement on local pairwise ranking decisions. We observe strong representational and moderate functional alignment within paradigms. Alignment across paradigms is low and task-dependent, yet stable across models. Our results suggest that the similar downstream performance observed across paradigms in prior work can be attributed to agreement on the relative importance of documents, rather than to agreement on rankings of the full training data.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2611.19286
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2611.19286 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2611.19286 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2611.19286 in a Space README.md to link it from this page.

Collections including this paper 1