Metriq PRISM Laboratory is an interdisciplinary research laboratory of Metriq Foundation, Inc. PRISM conducts exploratory research involving computational, mathematical, scientific, and machine-assisted methods.
A prism separates a difficult signal into components that can be inspected. PRISM applies the same principle to hard problems: define the object, isolate the unknowns, construct a method, expose the evidence, and publish the remaining uncertainty.
PRISM is designed to make technical and scientific work inspectable before anyone is asked to accept the conclusion.
PROGRAM FOR RESEARCH IN INTELLIGENT SYSTEMS AND METHODS
Define the claim. Expose the method. Publish the evidence and limitations.
PRISM develops candidate results through construction search, symbolic and numerical analysis, literature and prior-art review, proof development, exact computational checks, reproducible builds, and targeted adversarial review.
Search for explicit examples, counterexamples, algorithms, and proof structures that can be challenged directly.
Use symbolic, numerical, exact, and adversarial methods within clearly stated limits.
Release the paper, source, verification record, version history, and unresolved review targets together.
The catalog below reflects the canonical MetriqOrg/PRISM repository snapshot dated 18 July 2026. Six papers include interactive website explainers; the two newest papers point directly to their controlling GitHub records.
MF-PRISM-MATH; the former MF-MATH identifiers and releases remain preserved as historical predecessors. Metriq PRISM Laboratory is the organizational creator and issuer. Daniel H. Jeffery is Research Director and corresponding contributor. Paper 05 remains PUBLICATION HOLD · SOURCE-AUTHOR CLARIFICATION REQUIRED.
MF-PRISM-MATH-2026-01An explicit base-4 achievement-set construction presented as a candidate affirmative solution to the Second Jones Problem. The proposed sequence is strictly decreasing, sums to 17, has consecutive-term ratio converging to one half, and is argued to generate a Cantorval.
MF-PRISM-MATH-2026-02A variable-base achievement-set construction presented as a candidate affirmative solution to Problem 23(i): the existence of an achievable Cantorval whose boundary has Hausdorff dimension one.
MF-PRISM-MATH-2026-03A candidate proof that every achievement set generated by an absolutely convergent real series with infinitely many nonzero terms can be written as the Minkowski sum of two achievement sets of Hausdorff dimension zero.
MF-PRISM-MATH-2026-04A factorial mixed-radix construction presented as a candidate affirmative answer to whether an achievable Cantorval can have boundary of Hausdorff dimension zero.
MF-PRISM-MATH-2026-05An interpretation note analyzing the literal formulation of Problem 13 and a middle-third witness. Publication remains on hold pending clarification of the source authors' intended formulation.
MF-PRISM-MATH-2026-06Version 2.2 presents three exact finite subset-sum certificates for unresolved cardinal-function ranges, together with Cantor-set extensions and reproducible verification. The interactive explorer on this site visualizes the first two certificates; the canonical repository contains the complete current paper.
A candidate identity connecting endpoint walks on path graphs with complete homogeneous symmetric norms. The release includes exact computational consistency checks across 1,176 parameter pairs and a complete reproducibility package in the canonical repository.
A candidate proof of Mathar's conjectured recurrence for the Type-ace lattice-walk sequence OEIS A302186. The supplied exact verifier checks coefficients through n = 100, a degree-74 formal series, and the Ore-operator identity; the original verifier independently reruns through n = 80.
A speculative idea does not become an established claim through computational checks alone. PRISM separates development, reproducibility, and independent review.
Separate definitions, known results, constraints, unknowns, and exact proof obligations.
Translate intuition into explicit constructions, equations, assumptions, and failure conditions.
Develop candidate proofs, counterexamples, algorithms, or systems that can be tested directly.
Run exact checks, simulations, independent reproductions, and adversarial review within stated limits.
Release the manuscript, source, version history, review guide, and remaining uncertainty.
Generative artificial intelligence and other computational systems may be used as research instruments. Their use does not constitute independent verification. Metriq Foundation, acting through PRISM, determines what is published and accepts responsibility for issuing each work as a candidate result.
The repository distinguishes reproducibility materials from independent review and recognized scholarly peer review.
Published by PRISM but not independently peer reviewed.
Source, checks, and stated limits are provided for reproducibility.
Reviewed by an unaffiliated qualified specialist.
Completed a recognized external peer-review process.
Material revisions and superseded releases remain traceable.
MetriqOrg/PRISM is the authoritative public record for current papers, source materials, verification code, figures, checksums, revision histories, and reproducibility packages.
PUBLIC REPOSITORYCurrent files are never substituted silently for an earlier version. Corrections receive a new version number, and superseded material remains traceable through tagged releases, archival directories, or both.
git clone https://github.com/MetriqOrg/PRISM.gitOpen examination
PRISM expressly invites qualified criticism, independent replication, relevant prior art, corrections, and clearer arguments.