{"paper":{"title":"Sheaf-Theoretic Transport and Obstruction for Detecting Scientific Theory Shift in AI Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Sheaf obstruction measures rank the intended theory deformation or extension as the lowest-failure candidate in AI agent transitions.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"David N. Olivieri, Roque J. Hern\\'andez","submitted_at":"2026-05-13T18:46:17Z","abstract_excerpt":"Scientific theory shift in AI agents requires more than fitting equations to data. An artificial scientific agent must detect whether an existing representational framework remains transportable into a new regime, or whether its language has become locally-to-globally obstructed and must be extended. This paper develops a finite sheaf-theoretic framework for detecting theory-shift candidates through transport and obstruction. Contexts are organized as a local-to-global structure in which source, overlap, target, and validation charts are fitted, restricted, and tested for gluing. Obstruction m"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"The main result is direct obstruction ranking: the intended deformation or extension is usually the lowest-obstruction candidate, and transition type is separated in the benchmark.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the five obstruction components (residual fit, overlap incompatibility, constraint violation, limiting-relation failure, representational cost) can be defined and combined in a way that reliably separates deformation from extension without post-hoc tuning on the benchmark itself.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A finite sheaf-theoretic framework ranks obstruction measures to identify when an AI agent's theory must deform within its language or extend to a new one, validated on a controlled transition benchmark.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Sheaf obstruction measures rank the intended theory deformation or extension as the lowest-failure candidate in AI agent transitions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"46daafb31665dc01152ee40579a603bd024c3cd841612658cfc76ac269a9305d"},"source":{"id":"2605.14033","kind":"arxiv","version":1},"verdict":{"id":"3e734ad1-5591-4edb-beb2-1f60e473a701","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:41:38.313087Z","strongest_claim":"The main result is direct obstruction ranking: the intended deformation or extension is usually the lowest-obstruction candidate, and transition type is separated in the benchmark.","one_line_summary":"A finite sheaf-theoretic framework ranks obstruction measures to identify when an AI agent's theory must deform within its language or extend to a new one, validated on a controlled transition benchmark.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the five obstruction components (residual fit, overlap incompatibility, constraint violation, limiting-relation failure, representational cost) can be defined and combined in a way that reliably separates deformation from extension without post-hoc tuning on the benchmark itself.","pith_extraction_headline":"Sheaf obstruction measures rank the intended theory deformation or extension as the lowest-failure candidate in AI agent transitions."},"references":{"count":43,"sample":[{"doi":"10.1088/1367-2630/13/11/113036","year":2011,"title":"The sheaf- theoretic structure of non-locality and contextuality.New Journal of Physics, 13(11):113036, November 2011.doi: 10.1088/1367-2630/13/11/113036","work_id":"864e66a4-b792-40d0-92b7-7a2e152988f0","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"2025 Sheaf theory: from deep geometry to deep learning","work_id":"46192dc2-30c6-4b4b-b5fa-46af94536e45","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":", author Di Giovanni, F","work_id":"78a4a5a4-a86a-40ba-bc89-9de8a7932aae","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/icdm.2005.132","year":2005,"title":"author Borgwardt, K. M. , & author Kriegel, H.-P. ( year 2005 ). title Shortest-path kernels on graphs . In booktitle Proceedings of the Fifth IEEE International Conference on Data Mining \\/ (pp. page","work_id":"a67c0cae-76c1-4ee0-accb-50429a27e10e","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1073/pnas.1517384113","year":2016,"title":"Abhimanyu Das, Weihao Kong, Rajat Sen, and Yichen Zhou","work_id":"cbcc5fe8-a571-4a91-8ab8-1c0164648749","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":43,"snapshot_sha256":"e0b173b782b6166451f64ae7cca9cb720b71258094d39e076fa0c20f38b40e43","internal_anchors":2},"formal_canon":{"evidence_count":1,"snapshot_sha256":"1f9324a0f241e941928c7d3f04f77e1acfd689c3352f3339f06652a1252829d1"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}