Representation Fidelity:Auditing Algorithmic Decisions About Humans Using Self-Descriptions
This work addresses the problem of ensuring algorithmic decisions about humans are based on reasonable grounds, which is crucial for fairness and accountability in AI systems.
This paper introduces Representation Fidelity, a new metric to validate algorithmic decisions about humans by measuring the distance between an algorithm's input representation of a person and that person's self-description. They created the Loan-Granting Self-Representations Corpus 2025, a benchmark dataset of 30,000 synthetic natural language self-descriptions for loan-granting decisions, along with expert annotations of representation mismatches.
This paper introduces a new dimension for validating algorithmic decisions about humans by measuring the fidelity of their representations. Representation Fidelity measures if decisions about a person rest on reasonable grounds. We propose to operationalize this notion by measuring the distance between two representations of the same person: (1) an externally prescribed input representation on which the decision is based, and (2) a self-description provided by the human subject of the decision, used solely to validate the input representation. We examine the nature of discrepancies between these representations, how such discrepancies can be quantified, and derive a generic typology of representation mismatches that determine the degree of representation fidelity. We further present the first benchmark for evaluating representation fidelity based on a dataset of loan-granting decisions. Our Loan-Granting Self-Representations Corpus 2025 consists of a large corpus of 30 000 synthetic natural language self-descriptions derived from corresponding representations of applicants in the German Credit Dataset, along with expert annotations of representation mismatches between each pair of representations.