Utility-Aware Data Pricing: Token-Level Quality and Empirical Training Gain for LLMs

TL;DR

This paper introduces a utility-aware data valuation framework for LLMs that measures token-level information, empirical training gains, and cryptographic verifiability, enabling fair and transparent data pricing.

Contribution

It proposes a novel multi-layered approach combining information metrics, influence-based gain measurement, and cryptographic commitments for data valuation in LLMs.

Findings

Proxy-based empirical gain aligns closely with actual utility.

Outperforms traditional row-count and token-count baselines.

Validated on instruction, reasoning, and code tasks.

Abstract

Traditional data valuation methods based on ``row-count $\times$ quality coefficient'' paradigms fail to capture the nuanced, nonlinear contributions that data makes to Large Language Model (LLM) capabilities. This paper presents a dynamic data valuation framework that transitions from static accounting to utility-based pricing. Our approach operates on three layers: (1) token-level information density metrics using Shannon entropy and Data Quality Scores; (2) empirical training gain measurement through influence functions, proxy model strategies, and Data Shapley values; and (3) cryptographic verifiability through hash-based commitments, Merkle trees, and a tamper-evident training ledger. We provide comprehensive experimental validation on three real domains (instruction following, mathematical reasoning, and code summarization), demonstrating that proxy-based empirical gain achieves near-perfect ranking alignment with realized utility, substantially outperforming row-count and token-count baselines. This framework enables a fair Data-as-a-Service economy where high-reasoning data is priced according to its actual contribution to model intelligence, while providing the transparency and auditability necessary for trustworthy data markets.