{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:C6RBPSOPRL7MOKH5W2RT7IF6ZS","short_pith_number":"pith:C6RBPSOP","schema_version":"1.0","canonical_sha256":"17a217c9cf8afec728fdb6a33fa0becc844bdec93338f8328bf5548ab93d29c7","source":{"kind":"arxiv","id":"2607.18532","version":1},"attestation_state":"computed","paper":{"title":"Reasoning Fine-Tuning Induces Persistent Latent Policy States","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abir Harrasse, Chaithanya Bandi, Fateme Hashemi Chaleshtori, Hunar Batra, Michael Lan","submitted_at":"2026-07-20T21:56:22Z","abstract_excerpt":"Reasoning-specialized language models show large performance gains over base models, yet the internal changes responsible for improved multi-step reasoning remain poorly understood. It is unclear whether reasoning fine-tuning improves local token-level competence or globally reorganizes how models structure inference over time. We address this question by modeling Chain-of-Thought reasoning as a switching dynamical system (SDS), in which internal representations evolve under discrete latent policy states. Our framework combines time-aware contrastive representation learning with discrete regim"},"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":"2607.18532","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-20T21:56:22Z","cross_cats_sorted":[],"title_canon_sha256":"b4467e66542cd8fe72c645bd32ec334487592f7a3023751a774423028f1e9be1","abstract_canon_sha256":"4008e45b96e852cfad475628cc60dbbdea91606f3ffeb7c371fd2c97584d0fa7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-22T00:22:53.066392Z","signature_b64":"keeIGzqDwIxbRAQ8RtrapIN0IzJSzu+4AJhm+ra9Apsd/EWIrV2F9dT6yvPSwkpkMz7CoPeHDOjs1xpO0/bWBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"17a217c9cf8afec728fdb6a33fa0becc844bdec93338f8328bf5548ab93d29c7","last_reissued_at":"2026-07-22T00:22:53.065655Z","signature_status":"signed_v1","first_computed_at":"2026-07-22T00:22:53.065655Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reasoning Fine-Tuning Induces Persistent Latent Policy States","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abir Harrasse, Chaithanya Bandi, Fateme Hashemi Chaleshtori, Hunar Batra, Michael Lan","submitted_at":"2026-07-20T21:56:22Z","abstract_excerpt":"Reasoning-specialized language models show large performance gains over base models, yet the internal changes responsible for improved multi-step reasoning remain poorly understood. It is unclear whether reasoning fine-tuning improves local token-level competence or globally reorganizes how models structure inference over time. We address this question by modeling Chain-of-Thought reasoning as a switching dynamical system (SDS), in which internal representations evolve under discrete latent policy states. Our framework combines time-aware contrastive representation learning with discrete regim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.18532","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/2607.18532/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":"2607.18532","created_at":"2026-07-22T00:22:53.066063+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.18532v1","created_at":"2026-07-22T00:22:53.066063+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.18532","created_at":"2026-07-22T00:22:53.066063+00:00"},{"alias_kind":"pith_short_12","alias_value":"C6RBPSOPRL7M","created_at":"2026-07-22T00:22:53.066063+00:00"},{"alias_kind":"pith_short_16","alias_value":"C6RBPSOPRL7MOKH5","created_at":"2026-07-22T00:22:53.066063+00:00"},{"alias_kind":"pith_short_8","alias_value":"C6RBPSOP","created_at":"2026-07-22T00:22:53.066063+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/C6RBPSOPRL7MOKH5W2RT7IF6ZS","json":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS.json","graph_json":"https://pith.science/api/pith-number/C6RBPSOPRL7MOKH5W2RT7IF6ZS/graph.json","events_json":"https://pith.science/api/pith-number/C6RBPSOPRL7MOKH5W2RT7IF6ZS/events.json","paper":"https://pith.science/paper/C6RBPSOP"},"agent_actions":{"view_html":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS","download_json":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS.json","view_paper":"https://pith.science/paper/C6RBPSOP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.18532&json=true","fetch_graph":"https://pith.science/api/pith-number/C6RBPSOPRL7MOKH5W2RT7IF6ZS/graph.json","fetch_events":"https://pith.science/api/pith-number/C6RBPSOPRL7MOKH5W2RT7IF6ZS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS/action/storage_attestation","attest_author":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS/action/author_attestation","sign_citation":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS/action/citation_signature","submit_replication":"https://pith.science/pith/C6RBPSOPRL7MOKH5W2RT7IF6ZS/action/replication_record"}},"created_at":"2026-07-22T00:22:53.066063+00:00","updated_at":"2026-07-22T00:22:53.066063+00:00"}