{"paper":{"title":"SIGMA: Semantic Identifier Grouping for Molecular Autoregression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Fei Dou, Jinbo Bi, Minghu Song, Xinyu Wang","submitted_at":"2026-03-26T05:55:17Z","abstract_excerpt":"Autoregressive molecular models assign probability to molecular serializations even though chemical identity is invariant to serialization. Equivalent serializations can therefore represent a common molecular identity yet induce inconsistent next-token decisions. Randomized strings broaden exposure, but do not reveal which intermediate decisions should agree. We introduce SIGMA, a dense suffix-position objective built from chemically certified same-suffix triplets: two equivalent histories, one non-equivalent history, and a shared suffix. SIGMA aligns corresponding pre-token hidden states alon"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.25062","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.25062/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}