{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:O2IOEKP53JKJ54QPO2AGLKRXJM","short_pith_number":"pith:O2IOEKP5","schema_version":"1.0","canonical_sha256":"7690e229fdda549ef20f768065aa374b141705ea0248dc009902403b817a3ff5","source":{"kind":"arxiv","id":"2403.19827","version":3},"attestation_state":"computed","paper":{"title":"Language Models Learn Rare Phenomena from Less Rare Phenomena: The Case of the Missing AANNs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kanishka Misra, Kyle Mahowald","submitted_at":"2024-03-28T20:35:10Z","abstract_excerpt":"Language models learn rare syntactic phenomena, but the extent to which this is attributable to generalization vs. memorization is a major open question. To that end, we iteratively trained transformer language models on systematically manipulated corpora which were human-scale in size, and then evaluated their learning of a rare grammatical phenomenon: the English Article+Adjective+Numeral+Noun (AANN) construction (``a beautiful five days''). We compared how well this construction was learned on the default corpus relative to a counterfactual corpus in which AANN sentences were removed. We fo"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2403.19827","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T20:35:10Z","cross_cats_sorted":[],"title_canon_sha256":"3638283d80d395fd031501f90f9cda90e878ec21a493e5b62845d56c02f572d7","abstract_canon_sha256":"9a4b6e3b3fe7d47733de2c5ad62cb0741f440dae00e76cd6e8aec59b4b647d15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:26:42.232022Z","signature_b64":"8ZosaQbruLFyqy4QAysiW1+wt67lUgvUAnV+K5wBGMWuET27eV0aEQJapQ9a5zN7bOC1Y5WzKXR3aQUjORmQCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7690e229fdda549ef20f768065aa374b141705ea0248dc009902403b817a3ff5","last_reissued_at":"2026-07-05T11:26:42.231488Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:26:42.231488Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language Models Learn Rare Phenomena from Less Rare Phenomena: The Case of the Missing AANNs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kanishka Misra, Kyle Mahowald","submitted_at":"2024-03-28T20:35:10Z","abstract_excerpt":"Language models learn rare syntactic phenomena, but the extent to which this is attributable to generalization vs. memorization is a major open question. To that end, we iteratively trained transformer language models on systematically manipulated corpora which were human-scale in size, and then evaluated their learning of a rare grammatical phenomenon: the English Article+Adjective+Numeral+Noun (AANN) construction (``a beautiful five days''). We compared how well this construction was learned on the default corpus relative to a counterfactual corpus in which AANN sentences were removed. We fo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.19827","kind":"arxiv","version":3},"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/2403.19827/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2403.19827","created_at":"2026-07-05T11:26:42.231549+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.19827v3","created_at":"2026-07-05T11:26:42.231549+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.19827","created_at":"2026-07-05T11:26:42.231549+00:00"},{"alias_kind":"pith_short_12","alias_value":"O2IOEKP53JKJ","created_at":"2026-07-05T11:26:42.231549+00:00"},{"alias_kind":"pith_short_16","alias_value":"O2IOEKP53JKJ54QP","created_at":"2026-07-05T11:26:42.231549+00:00"},{"alias_kind":"pith_short_8","alias_value":"O2IOEKP5","created_at":"2026-07-05T11:26:42.231549+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26050","citing_title":"Natural Ungrokking: Asymmetric Control of Which Rules Survive Pretraining","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21645","citing_title":"Behavioral and Representational Evidence of Binomial Ordering Preferences in Large Language Models","ref_index":275,"is_internal_anchor":false},{"citing_arxiv_id":"2502.20349","citing_title":"Naturalistic Computational Cognitive Science: Towards generalizable models and theories that capture the full range of natural behavior","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23032","citing_title":"Brain-LLM Alignment Tracks Training Data, Not Typology","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23035","citing_title":"Sparse Autoencoders Map Brain-LLM Alignment onto Cortical Semantic Topography","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2408.05086","citing_title":"A systematic framework for generating novel experimental hypotheses from language models","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2502.20349","citing_title":"Naturalistic Computational Cognitive Science: Towards generalizable models and theories that capture the full range of natural behavior","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM","json":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM.json","graph_json":"https://pith.science/api/pith-number/O2IOEKP53JKJ54QPO2AGLKRXJM/graph.json","events_json":"https://pith.science/api/pith-number/O2IOEKP53JKJ54QPO2AGLKRXJM/events.json","paper":"https://pith.science/paper/O2IOEKP5"},"agent_actions":{"view_html":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM","download_json":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM.json","view_paper":"https://pith.science/paper/O2IOEKP5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.19827&json=true","fetch_graph":"https://pith.science/api/pith-number/O2IOEKP53JKJ54QPO2AGLKRXJM/graph.json","fetch_events":"https://pith.science/api/pith-number/O2IOEKP53JKJ54QPO2AGLKRXJM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM/action/storage_attestation","attest_author":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM/action/author_attestation","sign_citation":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM/action/citation_signature","submit_replication":"https://pith.science/pith/O2IOEKP53JKJ54QPO2AGLKRXJM/action/replication_record"}},"created_at":"2026-07-05T11:26:42.231549+00:00","updated_at":"2026-07-05T11:26:42.231549+00:00"}