Flood and Harvest: The Provable Necessity of Trivia for Generating Valuable Mathematics via the Lens of Language Generation in the Limit

2026-06-12

AI Summary

This work addresses the fundamental challenge of distinguishing verifiable yet trivial statements from genuinely valuable mathematical content in formalized theorem generation—a key step toward bridging the gap between automated provers’ capabilities and mathematicians’ value judgments. The authors model the generation of valuable mathematics as a nested language process in the limit, leveraging membership oracles for an unknown target language of valuable statements within a formal language framework, and introduce adversarial enumeration and density analysis techniques. Their central contribution reveals that coverage performance depends on the absolute quantity, not the density, of trivial statements: finitely many trivialities cover at most α/2 of the space, whereas infinitely many—even with vanishing density—can achieve coverage up to 1−α/2. They rigorously prove that no perfect verifier can substitute for mathematical taste and construct a generator attaining this dual coverage bound.

Abstract

AI systems coupled to proof assistants now generate formal mathematics at scale, and the gap between what a checker can verify and what a mathematician would value has become the binding constraint. We model the generation of valuable mathematics as nested language generation in the limit: a verifiable formal language $F$, accessed through a membership oracle (the proof checker), contains an unknown valuable language $H \in \mathcal{H}$ revealed only through an adversarial enumeration of a core $C \subseteq H$ of exact density $α$ (the literature). Every output is valuable ($\in H$), trivial ($\in F \setminus H$), or a hallucination ($\notin F$). We settle four questions. First, the verifier is not taste: the collections admitting generation with breadth are exactly those of the oracle-free model, characterized fiber-wise by Angluin's condition. Second, the verifier does buy sound coverage, covering all unseen valuable statements while asserting only valid ones: possible with it, impossible without it; it relocates unavoidable errors from false to trivial. Third, and centrally, a sharp dichotomy on the tight family: generators emitting finitely many trivia achieve optimal coverage $α/2$, while any infinite trivia allowance, even at vanishing rate, jumps the optimum to $1-α/2$ (both tight, for cores presented as the candidate intersection), and one generator attains both ends. The transition is in trivia count, not rate; the gap $1-α$ is the unrecorded mass. Fourth, both regimes instantiate in a compression model of mathematics. A perfect verifier cannot substitute for taste: the unbounded stream of correct-but-worthless statements is not an engineering accident but a provable necessity, since covering unrecorded valuable mathematics requires an infinite, but asymptotically negligible, stream of certified trivia.

Problem

• formal mathematics
• value
• verifiability
• language generation
• trivia

Innovation

• language generation in the limit
• formal mathematics generation
• trivia necessity
• verifier vs taste
• coverage dichotomy