Answers are grounded in this paper and cite their sources.
This comprehensive archive details the mass-gap resolution process in Yang-Mills theory, utilizing AI tools and mathematical frameworks.
to be honest, the mass gap was as boring as poincare and so we lost interest in closing the ustc, there is no incentive nor pressure to finish it, and due to llm limitations, we will leave it open as a candyThis Zenodo deposit preserves the complete working archive of the MassGap-Carré project. The archive documents the development of a proposed complete ZFC resolution of the Yang–Mills mass-gap problem, together with the calculations, reconstructions, corrections, conversations, intermediate assemblies, verification attempts, discarded branches, and auxiliary confinement work produced during the research process. The deposit is intentionally broader than a conventional final paper. It is presented as a making-of archive: a record of how the solution was discovered, calculated, formalized, challenged, reconstructed, and progressively assembled. The organization of the files reflects the actual history of the work. It is therefore not perfectly linear. Some documents overlap, some revisit earlier calculations, some preserve alternative formulations, and some contain conversations or work sessions rather than polished mathematical exposition. The principal families of documents are described below. The Ligne droite series The Ligne droite series contains the earliest direct calculations leading toward the complete ZFC mass-gap construction. These documents record the first continuous attempt to move directly through the problem without separating the research into conventional papers or predetermined formal stages. They contain the initial development of the main mathematical route, including the finite gauge-theoretic framework, spectral constructions, multiscale mechanisms, positivity structures, continuum questions, and the successive obstacles encountered during the attempt to obtain a complete proof. The series should be read as the first direct computational backbone of the project. It preserves: the earliest forms of the proposed solution; direct calculations made before the final terminology was stabilized; intermediate definitions and discarded formulations; the first appearance of several structures later reorganized in the formal manuscripts; the chronological mathematical path by which the mass-gap rail was discovered. The Ligne droite documents are therefore not merely drafts of the final manuscript. They are the primary research trace of the original direct calculation. The Calculations series The Calculations series contains the formal mathematical work undertaken to convert the direct Ligne droite construction into an explicit ZFC proof architecture. These files are largely preserved as copied working conversations. They document, in real time, the use of large language models as mathematical collaborators, calculation assistants, critics, reconstruction tools, and formalization engines. Their purpose was to take the direct calculations and require that every relevant object be: explicitly defined; placed in a finite or countable ZFC framework; recalculated rather than accepted through informal intuition; tested for hidden assumptions; connected to the next stage of the argument; rewritten in theorem–lemma–proof form; checked against possible logical gaps; made suitable for integration into a complete proof chain. The Calculations series therefore preserves both the mathematical content and the process by which LLM systems were used to formalize it. These conversations are included directly because they show the effective role of LLMs in the development of the solution. They record not only polished answers, but also corrections, disagreements, failed interpretations, repeated calculations, source recovery, debugging, and the gradual production of the formal structure. They are part of the provenance of the research. The Productions series The Productions series contains the later recalculations, reorganizations, proof assemblies, and manuscript-building work. After the direct route and formal calculation phases, the project required repeated reconstruction of the proof into larger coherent blocks. The Productions documents record this phase. They include, in varying degrees of completion: recalculated versions of earlier arguments; consolidation of definitions and proof dependencies; reconciliation of parallel notations; staged assemblies of the spectral rail; attempts to produce self-contained mathematical documents; verification and correction of previously generated material; preparation of complete or partial manuscripts; integration of auxiliary calculations; construction of final proof packets. Like the Calculations series, many Production files are direct copies of LLM-assisted conversations. They show how the mathematical material was recomputed and assembled through repeated interaction with language models. The Productions series is therefore both a mathematical record and a record of the practical process of building a long proof with contemporary AI systems. Mass-gap component The central body of the archive concerns the proposed strict (SU(3)) spectral mass-gap construction. The final written rail is organized around a progression of finite gauge-theoretic and Osterwalder–Schrader constructions, produced/true comparisons, multiscale controls, continuum passage, observable nontriviality, spectral assembly, and local-field completion. The principal claimed conclusion is [ H_∞YM CΩ,] [Spec(H_∞YM)⊂{0}∪[mcert,∞),] with [mcert 4κ/3ΞJ14>0,] and consequently [mphyscert>0.] The archive includes the calculations and successive written forms used to construct, explain, assemble, and verify this claimed result. Rather than asking readers to rely only on a final polished document, the deposit makes the development record available so that the origin of the formulas, proof choices, corrections, and dependencies can be independently examined. USTC and the confinement firewall The archive also contains a later research branch concerning Wilson area-law and confinement questions, referred to internally as USTC. This branch was developed separately from the spectral mass-gap argument. Its principal completed role is that of a logical firewall. It formulates the precise positive-witness criterion required to pass from local or cellwise estimates to a Wilson area-law conclusion, while preventing unproved confinement claims from entering the spectral mass-gap proof. The intended boundary is expressed by statements of the form [POSITIVE_WITNESS_GATE=CLOSED_ZFC,] [POSITIVE_WITNESS\ NOT\ EMITTED,] [USTC_CLOSED_ZFC\ NOT\ CLAIMED,] and [WILSON_AREA_LAW CONDITIONAL_ON_POSITIVE_USTC_WITNESS.] Accordingly, the USTC material should be read primarily as a confinement criterion, a firewall, and a record of the attempted extension beyond the spectral result. It is not used as a premise of the mass-gap rail. Why further USTC work was discontinued Further work on the USTC branch was discontinued after the author lost interest in continuing that extension under the existing LLM-assisted workflow. The principal difficulty was no longer the production of individual mathematical calculations. It was the inability of current large language models to maintain a stable understanding of a very large, long-running project. During repeated sessions, the systems frequently: lost the active state of the project; treated successive tasks as unrelated projects; reopened stages that had already been completed; replaced assembly work with new audit programs; generated parallel folders and incompatible manuscript branches; confused verification of source transfer with mathematical proof; forgot which results were inputs and which were still being produced; repeated work that had already been completed; redirected the project away from the requested task; consumed substantial time reconstructing context rather than advancing the mathematics. The resulting problem became one of workflow instability, context fragmentation, duplicated verification, and unreliable long-horizon continuity. For this reason, the USTC continuation no longer justified the time required. The existing USTC materials are nevertheless included so that the work is preserved and so that any later researcher, verification system, or more capable AI tool can continue from the actual archive rather than restarting the project. LLM-assisted research record A central purpose of this deposit is to document the role of large language models in the development of the MassGap-Carré project. The archive does not hide the conversational origin of many calculations. The Calculations and Productions series include direct copied conversations because those conversations record: requests for explicit mathematical calculations; derivation of missing formulas; formalization of informal arguments; detection of contradictions; reconstruction of incomplete proof blocks; generation of finite tables and ledgers; criticism of previous answers; correction of notation and scope; assembly of large mathematical documents; repeated attempts to verify ZFC completeness. The archive therefore provides material for studying both the potential and the limitations of LLM-assisted mathematical research. It shows that language models can contribute to a large mathematical construction, but also that they can become the principal source of delay when the project exceeds their reliable context-management capacity. Nature of the archive This deposit should be understood as: a complete research-development archive; a record of a proposed ZFC mass-gap resolut
No takes yet. Share an insight, caveat, or question.
Son David Bolduc (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: