The Structure of Relation Decoding Linear Operators in Large Language Models
2510.26543v1
cs.CL, cs.AI, cs.LG
2025-11-01
Авторы:
Miranda Anna Christ, Adrián Csiszárik, Gergely Becsó, Dániel Varga
Abstract
This paper investigates the structure of linear operators introduced in
Hernandez et al. [2023] that decode specific relational facts in transformer
language models. We extend their single-relation findings to a collection of
relations and systematically chart their organization. We show that such
collections of relation decoders can be highly compressed by simple order-3
tensor networks without significant loss in decoding accuracy. To explain this
surprising redundancy, we develop a cross-evaluation protocol, in which we
apply each linear decoder operator to the subjects of every other relation. Our
results reveal that these linear maps do not encode distinct relations, but
extract recurring, coarse-grained semantic properties (e.g., country of capital
city and country of food are both in the country-of-X property). This
property-centric structure clarifies both the operators' compressibility and
highlights why they generalize only to new relations that are semantically
close. Our findings thus interpret linear relational decoding in transformer
language models as primarily property-based, rather than relation-specific.
Ссылки и действия
Дополнительные ресурсы: