Paper: arXiv 2506.07066

Authors: Li Jingyuan

Abstract

This paper presents a comprehensive formalization of the von Neumann-Morgenstern (vNM) expected utility theorem using the Lean 4 interactive theorem prover. We implement the classical axioms of preference-completeness, transitivity, continuity, and independence-enabling machine-verified proofs of both the existence and uniqueness of utility representations. Our formalization captures the mathematical structure of preference relations over lotteries, verifying that preferences satisfying the vNM axioms can be represented by expected utility maximization. Our contributions include a granular implementation of the independence axiom, formally verified proofs of fundamental claims about mixture lotteries, constructive demonstrations of utility existence, and computational experiments validating the results. We prove equivalence to classical presentations while offering greater precision at decision boundaries. This formalization provides a rigorous foundation for applications in economic modeling, AI alignment, and management decision systems, bridging the gap between theoretical decision theory and computational implementation.

Complexity vs Empirical Score

  • Math Complexity: 9.0/10
  • Empirical Rigor: 2.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper relies heavily on advanced formal mathematics, including measure theory and dependent type theory for axiomatization in Lean 4, which pushes the math complexity to the upper end of the scale. However, it lacks real-world data or backtesting, focusing instead on theoretical verification, resulting in low empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Formalize vNM Utility Theorem<br>in Lean 4"] --> B["Methodology: Axioms<br>Completeness, Transitivity,<br>Continuity, Independence"]
  B --> C["Input: Lotteries<br>Preference Relations"]
  C --> D["Computational Process<br>Mechanized Proofs<br>Mixture Properties"]
  D --> E["Outcome: Verified Theorem<br>Existence & Uniqueness<br>of Utility Representation"]
  E --> F["Applications<br>Economic Modeling, AI Alignment,<br>Decision Systems"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a comprehensive formalization and mechanized proof of the von Neumann‑Morgenstern utility theorem using the Lean 4 interactive theorem prover.

Source sentence
This paper presents a comprehensive formalization of the von Neumann-Morgenstern (vNM) expected utility theorem using the Lean 4 interactive theorem prover.

Assumes 2

  • Preferences satisfy the vNM axioms of completeness, transitivity, continuity, and independence. model
    Source sentence
    We implement the classical axioms of preference-completeness, transitivity, continuity, and independence-enabling machine-verified proofs of both the existence and uniqueness of utility representations.
  • A preference dataset is non‑empty and contains no direct contradictions (no cycles). data
    Source sentence
    PrefDataset (X : Type) [Fintype X] [Nonempty X] [DecidableEq X] where 11 /-- List of preference pairs (p \succ q) -/ 12 pairs : List (Lottery X \times Lottery X) 13 14 /- Assumption: under certain conditions on the dataset, 15 the learned reward model satisfies VNM axioms -/ 16 /-- Checks if dataset has sufficient coverage of preference space -/ 17 def datasetCoverage (data : PrefDataset X) : Prop := 18

Data

  • PrefDataset synthetic unknown
    Source sentence
    PrefDataset (X : Type) [Fintype X] [Nonempty X] [DecidableEq X] where 11 /-- List of preference pairs (p \succ q) -/ 12 pairs : List (Lottery X \times Lottery X) 13 14 /- Assumption: under certain conditions on the dataset, 15 the learned reward model satisfies VNM axioms -/ 16 /-- Checks if dataset has sufficient coverage of preference space -/ 17 def datasetCoverage (data : PrefDataset X) : Prop := 18

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • The formalization captures the mathematical structure of preference relations over lotteries and verifies that preferences satisfying the vNM axioms can be represented by expected utility maximization. theoretical
    Source sentence
    Our formalization captures the mathematical structure of preference relations over lotteries, verifying that preferences satisfying the vNM axioms can be represented by expected utility maximization.
  • The mechanized proof establishes equivalence to classical presentations while offering greater precision at decision boundaries. theoretical
    Source sentence
    We prove equivalence to classical presentations while offering greater precision at decision boundaries.
  • Executable implementations of utility representations enable computational experiments and empirical validation of the formalized results. theoretical
    Source sentence
    Our contributions include a granular implementation of the independence axiom, formally verified proofs of fundamental claims about mixture lotteries, constructive demonstrations of utility existence, and computational experiments validating the results. A complete, mechanized proof of the vNM utility theorem with explicit, precisely formulated axioms 2.
  • The work demonstrates that formal verification tools can successfully capture the nuanced mathematical structure of foundational economic results. theoretical
    Source sentence
    Our work demonstrates that formal verification tools can successfully capture the nuanced mathematical structure of foundational economic results.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 12 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

https://github.com/jingyuanli-hk/vNM-Theorem-pub — the authors' own repository

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.