{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UYZGJIRR2U3DV7SK6TEMTAVQZ7","short_pith_number":"pith:UYZGJIRR","schema_version":"1.0","canonical_sha256":"a63264a231d5363afe4af4c8c982b0cfe983456eff852b34204499876b0d55f5","source":{"kind":"arxiv","id":"2606.20937","version":1},"attestation_state":"computed","paper":{"title":"Learning through Internalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Julia Kempe, Marko Medvedev, Nati Srebro, Nikolaos Tsilivis, Nirmit Joshi","submitted_at":"2026-06-18T21:01:42Z","abstract_excerpt":"We study internalization processes, by which neural-network-based systems absorb an explicit computational procedure into their own weights, and how they facilitate learning. We investigate how transformers internalize the simulation of semiautomata by internalizing chain-of-thought (CoT) tokens, which classes of semiautomata are harder to internalize, and expose the flip side of internalization, that is, a progressive degradation of out-of-distribution performance. We then provide the first provable analysis of successful internalization: for the task of learning parities, we show that a simp"},"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":"2606.20937","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-18T21:01:42Z","cross_cats_sorted":[],"title_canon_sha256":"56acbba83a09226a21dd0b1127189220114a23162e6aa64fb794ca38ecaa7bb0","abstract_canon_sha256":"4d90cc480ed223e5860f3dbbc52d884a12389117bcb528615b4f5ef6d94e057e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T01:12:22.581094Z","signature_b64":"LstZanV43Bhg3oyo3spU/OBA6AFkIc2bhc7SMTGbjYaO0c9AeLtVBZ907AeV+AX6+oKOB4KWDJqAzU5OEb3xCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a63264a231d5363afe4af4c8c982b0cfe983456eff852b34204499876b0d55f5","last_reissued_at":"2026-06-23T01:12:22.580621Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T01:12:22.580621Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning through Internalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Julia Kempe, Marko Medvedev, Nati Srebro, Nikolaos Tsilivis, Nirmit Joshi","submitted_at":"2026-06-18T21:01:42Z","abstract_excerpt":"We study internalization processes, by which neural-network-based systems absorb an explicit computational procedure into their own weights, and how they facilitate learning. We investigate how transformers internalize the simulation of semiautomata by internalizing chain-of-thought (CoT) tokens, which classes of semiautomata are harder to internalize, and expose the flip side of internalization, that is, a progressive degradation of out-of-distribution performance. We then provide the first provable analysis of successful internalization: for the task of learning parities, we show that a simp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20937","kind":"arxiv","version":1},"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/2606.20937/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":"2606.20937","created_at":"2026-06-23T01:12:22.580704+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.20937v1","created_at":"2026-06-23T01:12:22.580704+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20937","created_at":"2026-06-23T01:12:22.580704+00:00"},{"alias_kind":"pith_short_12","alias_value":"UYZGJIRR2U3D","created_at":"2026-06-23T01:12:22.580704+00:00"},{"alias_kind":"pith_short_16","alias_value":"UYZGJIRR2U3DV7SK","created_at":"2026-06-23T01:12:22.580704+00:00"},{"alias_kind":"pith_short_8","alias_value":"UYZGJIRR","created_at":"2026-06-23T01:12:22.580704+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7","json":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7.json","graph_json":"https://pith.science/api/pith-number/UYZGJIRR2U3DV7SK6TEMTAVQZ7/graph.json","events_json":"https://pith.science/api/pith-number/UYZGJIRR2U3DV7SK6TEMTAVQZ7/events.json","paper":"https://pith.science/paper/UYZGJIRR"},"agent_actions":{"view_html":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7","download_json":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7.json","view_paper":"https://pith.science/paper/UYZGJIRR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.20937&json=true","fetch_graph":"https://pith.science/api/pith-number/UYZGJIRR2U3DV7SK6TEMTAVQZ7/graph.json","fetch_events":"https://pith.science/api/pith-number/UYZGJIRR2U3DV7SK6TEMTAVQZ7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7/action/storage_attestation","attest_author":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7/action/author_attestation","sign_citation":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7/action/citation_signature","submit_replication":"https://pith.science/pith/UYZGJIRR2U3DV7SK6TEMTAVQZ7/action/replication_record"}},"created_at":"2026-06-23T01:12:22.580704+00:00","updated_at":"2026-06-23T01:12:22.580704+00:00"}