{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SM7F4YBXISVWMP277GB7XZMO7C","short_pith_number":"pith:SM7F4YBX","schema_version":"1.0","canonical_sha256":"933e5e603744ab663f5ff983fbe58ef8903e1bedf52c86a9bb007fe358ce5d59","source":{"kind":"arxiv","id":"2509.01432","version":1},"attestation_state":"computed","paper":{"title":"The Geometry of Nonlinear Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Nico Scherf, Nikola Milosevic","submitted_at":"2025-09-01T12:42:43Z","abstract_excerpt":"Reward maximization, safe exploration, and intrinsic motivation are often studied as separate objectives in reinforcement learning (RL). We present a unified geometric framework, that views these goals as instances of a single optimization problem on the space of achievable long-term behavior in an environment. Within this framework, classical methods such as policy mirror descent, natural policy gradient, and trust-region algorithms naturally generalize to nonlinear utilities and convex constraints. We illustrate how this perspective captures robustness, safety, exploration, and diversity obj"},"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":"2509.01432","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-09-01T12:42:43Z","cross_cats_sorted":[],"title_canon_sha256":"dfaf80a3d59fedf4c83fac0fe17f1cf65435981e1442510f3bdf6244fa9aeb42","abstract_canon_sha256":"1c7ae4474741785a1730f20e2d81203178821f09cbd699bba3d4bd8484f650b9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:00.849695Z","signature_b64":"t/8WeixLcY7PpGnL/+yGWHFoLdTrNudoXAtfhY61xXRBPIW9Glta3lWucZuNS+dAzThdC79o8kpj5Z9u9k+yAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"933e5e603744ab663f5ff983fbe58ef8903e1bedf52c86a9bb007fe358ce5d59","last_reissued_at":"2026-07-05T12:03:00.849197Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:00.849197Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Geometry of Nonlinear Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Nico Scherf, Nikola Milosevic","submitted_at":"2025-09-01T12:42:43Z","abstract_excerpt":"Reward maximization, safe exploration, and intrinsic motivation are often studied as separate objectives in reinforcement learning (RL). We present a unified geometric framework, that views these goals as instances of a single optimization problem on the space of achievable long-term behavior in an environment. Within this framework, classical methods such as policy mirror descent, natural policy gradient, and trust-region algorithms naturally generalize to nonlinear utilities and convex constraints. We illustrate how this perspective captures robustness, safety, exploration, and diversity obj"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.01432","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/2509.01432/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":"2509.01432","created_at":"2026-07-05T12:03:00.849261+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.01432v1","created_at":"2026-07-05T12:03:00.849261+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.01432","created_at":"2026-07-05T12:03:00.849261+00:00"},{"alias_kind":"pith_short_12","alias_value":"SM7F4YBXISVW","created_at":"2026-07-05T12:03:00.849261+00:00"},{"alias_kind":"pith_short_16","alias_value":"SM7F4YBXISVWMP27","created_at":"2026-07-05T12:03:00.849261+00:00"},{"alias_kind":"pith_short_8","alias_value":"SM7F4YBX","created_at":"2026-07-05T12:03:00.849261+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29092","citing_title":"Priced Motion Through Optimal Faces: A Normal-Fan Geometry for Non-Stationary Adversarial MDPs","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C","json":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C.json","graph_json":"https://pith.science/api/pith-number/SM7F4YBXISVWMP277GB7XZMO7C/graph.json","events_json":"https://pith.science/api/pith-number/SM7F4YBXISVWMP277GB7XZMO7C/events.json","paper":"https://pith.science/paper/SM7F4YBX"},"agent_actions":{"view_html":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C","download_json":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C.json","view_paper":"https://pith.science/paper/SM7F4YBX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.01432&json=true","fetch_graph":"https://pith.science/api/pith-number/SM7F4YBXISVWMP277GB7XZMO7C/graph.json","fetch_events":"https://pith.science/api/pith-number/SM7F4YBXISVWMP277GB7XZMO7C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C/action/storage_attestation","attest_author":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C/action/author_attestation","sign_citation":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C/action/citation_signature","submit_replication":"https://pith.science/pith/SM7F4YBXISVWMP277GB7XZMO7C/action/replication_record"}},"created_at":"2026-07-05T12:03:00.849261+00:00","updated_at":"2026-07-05T12:03:00.849261+00:00"}