{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:RSSSKIVGZZEU56OY6HMTCNZDYY","short_pith_number":"pith:RSSSKIVG","schema_version":"1.0","canonical_sha256":"8ca52522a6ce494ef9d8f1d9313723c638a3b82dc747fe08cb9eaab926717c90","source":{"kind":"arxiv","id":"2606.05201","version":1},"attestation_state":"computed","paper":{"title":"State commitment learning: training language models to distinguish computation from memory","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Fei Ding, Huiming Yang, Runhao Liu, Yongkang Zhang, Yuhao Liao, Zijian Zeng","submitted_at":"2026-05-22T17:05:31Z","abstract_excerpt":"Reasoning language models do not distinguish tokens used for computation from tokens that constitute persistent state: once generated, all hidden thoughts remain in context and influence future predictions. As a result, downstream reasoning may depend on failed attempts, dead ends, and private scratch work that should not be safely relied on later. We recast this phenomenon as a new training objective, state commitment learning: training models to explicitly distinguish information that should be committed as persistent state from temporary computation that can be discarded. We define a counte"},"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":"2606.05201","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-22T17:05:31Z","cross_cats_sorted":[],"title_canon_sha256":"60adf2d530c2a2ec6a74b2f58ec1aec2144e151b432059f0c3d3b949e5ef9762","abstract_canon_sha256":"16a78510b62ec7635291d62b26b94f76b96834aa178600ca3ee07fed51bc43df"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T00:13:48.569604Z","signature_b64":"f+ucLCtWD1PtYCjMcj/P3SR9LwicrrRH8myDjGTSHbJ+EgnWKGVdjf8FWERSUAiL+dlXNb0gkzuPy12a1ObyBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ca52522a6ce494ef9d8f1d9313723c638a3b82dc747fe08cb9eaab926717c90","last_reissued_at":"2026-06-05T00:13:48.569087Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T00:13:48.569087Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"State commitment learning: training language models to distinguish computation from memory","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Fei Ding, Huiming Yang, Runhao Liu, Yongkang Zhang, Yuhao Liao, Zijian Zeng","submitted_at":"2026-05-22T17:05:31Z","abstract_excerpt":"Reasoning language models do not distinguish tokens used for computation from tokens that constitute persistent state: once generated, all hidden thoughts remain in context and influence future predictions. As a result, downstream reasoning may depend on failed attempts, dead ends, and private scratch work that should not be safely relied on later. We recast this phenomenon as a new training objective, state commitment learning: training models to explicitly distinguish information that should be committed as persistent state from temporary computation that can be discarded. We define a counte"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05201","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/2606.05201/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":"2606.05201","created_at":"2026-06-05T00:13:48.569147+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.05201v1","created_at":"2026-06-05T00:13:48.569147+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05201","created_at":"2026-06-05T00:13:48.569147+00:00"},{"alias_kind":"pith_short_12","alias_value":"RSSSKIVGZZEU","created_at":"2026-06-05T00:13:48.569147+00:00"},{"alias_kind":"pith_short_16","alias_value":"RSSSKIVGZZEU56OY","created_at":"2026-06-05T00:13:48.569147+00:00"},{"alias_kind":"pith_short_8","alias_value":"RSSSKIVG","created_at":"2026-06-05T00:13:48.569147+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/RSSSKIVGZZEU56OY6HMTCNZDYY","json":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY.json","graph_json":"https://pith.science/api/pith-number/RSSSKIVGZZEU56OY6HMTCNZDYY/graph.json","events_json":"https://pith.science/api/pith-number/RSSSKIVGZZEU56OY6HMTCNZDYY/events.json","paper":"https://pith.science/paper/RSSSKIVG"},"agent_actions":{"view_html":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY","download_json":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY.json","view_paper":"https://pith.science/paper/RSSSKIVG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.05201&json=true","fetch_graph":"https://pith.science/api/pith-number/RSSSKIVGZZEU56OY6HMTCNZDYY/graph.json","fetch_events":"https://pith.science/api/pith-number/RSSSKIVGZZEU56OY6HMTCNZDYY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY/action/storage_attestation","attest_author":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY/action/author_attestation","sign_citation":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY/action/citation_signature","submit_replication":"https://pith.science/pith/RSSSKIVGZZEU56OY6HMTCNZDYY/action/replication_record"}},"created_at":"2026-06-05T00:13:48.569147+00:00","updated_at":"2026-06-05T00:13:48.569147+00:00"}