{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:G56WN5Q6KVSZFVMVRTCD5RQY5Z","short_pith_number":"pith:G56WN5Q6","schema_version":"1.0","canonical_sha256":"377d66f61e556592d5958cc43ec618ee4fec428f3d74bcd92372b628b61e2b5a","source":{"kind":"arxiv","id":"2509.03790","version":2},"attestation_state":"computed","paper":{"title":"What Fundamental Structure in Reward Functions Enables Efficient Sparse-Reward Learning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Anuj Sharma, Ibne Farabi Shihab, Sanjeda Akter","submitted_at":"2025-09-04T00:53:02Z","abstract_excerpt":"Sparse-reward reinforcement learning (RL) remains fundamentally hard: without structure, any agent needs $\\Omega(|\\mathcal{S}||\\mathcal{A}|/p)$ samples to recover rewards. We introduce Policy-Aware Matrix Completion (PAMC) as a first concrete step toward a structural reward learning framework. Our key idea is to exploit approximate low-rank + sparse structure in the reward matrix, under policy-biased (MNAR) sampling. We prove recovery guarantees with inverse-propensity weighting, and establish a visitation-weighted error-to-regret bound linking completion error to control performance. Importan"},"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":"2509.03790","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-04T00:53:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"502cc7e8a1d4c768290b94841560a879575058e89e9af8587409fe605010b82e","abstract_canon_sha256":"632b9c4e0e414428d422e683223e0374c6afee63862ffa7e7311389f6accbdb6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:07:11.699404Z","signature_b64":"NE+wAVQ/iMczk/1MmwhxfJCWV5jwHC6NbUMLXpo1JbCQBZPsjctGKDQmUnpUfv7h5YzyL1F2vR8RWMsskHa/CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"377d66f61e556592d5958cc43ec618ee4fec428f3d74bcd92372b628b61e2b5a","last_reissued_at":"2026-07-05T12:07:11.698978Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:07:11.698978Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Fundamental Structure in Reward Functions Enables Efficient Sparse-Reward Learning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Anuj Sharma, Ibne Farabi Shihab, Sanjeda Akter","submitted_at":"2025-09-04T00:53:02Z","abstract_excerpt":"Sparse-reward reinforcement learning (RL) remains fundamentally hard: without structure, any agent needs $\\Omega(|\\mathcal{S}||\\mathcal{A}|/p)$ samples to recover rewards. We introduce Policy-Aware Matrix Completion (PAMC) as a first concrete step toward a structural reward learning framework. Our key idea is to exploit approximate low-rank + sparse structure in the reward matrix, under policy-biased (MNAR) sampling. We prove recovery guarantees with inverse-propensity weighting, and establish a visitation-weighted error-to-regret bound linking completion error to control performance. Importan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.03790","kind":"arxiv","version":2},"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/2509.03790/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":"2509.03790","created_at":"2026-07-05T12:07:11.699034+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.03790v2","created_at":"2026-07-05T12:07:11.699034+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.03790","created_at":"2026-07-05T12:07:11.699034+00:00"},{"alias_kind":"pith_short_12","alias_value":"G56WN5Q6KVSZ","created_at":"2026-07-05T12:07:11.699034+00:00"},{"alias_kind":"pith_short_16","alias_value":"G56WN5Q6KVSZFVMV","created_at":"2026-07-05T12:07:11.699034+00:00"},{"alias_kind":"pith_short_8","alias_value":"G56WN5Q6","created_at":"2026-07-05T12:07:11.699034+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/G56WN5Q6KVSZFVMVRTCD5RQY5Z","json":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z.json","graph_json":"https://pith.science/api/pith-number/G56WN5Q6KVSZFVMVRTCD5RQY5Z/graph.json","events_json":"https://pith.science/api/pith-number/G56WN5Q6KVSZFVMVRTCD5RQY5Z/events.json","paper":"https://pith.science/paper/G56WN5Q6"},"agent_actions":{"view_html":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z","download_json":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z.json","view_paper":"https://pith.science/paper/G56WN5Q6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.03790&json=true","fetch_graph":"https://pith.science/api/pith-number/G56WN5Q6KVSZFVMVRTCD5RQY5Z/graph.json","fetch_events":"https://pith.science/api/pith-number/G56WN5Q6KVSZFVMVRTCD5RQY5Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z/action/storage_attestation","attest_author":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z/action/author_attestation","sign_citation":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z/action/citation_signature","submit_replication":"https://pith.science/pith/G56WN5Q6KVSZFVMVRTCD5RQY5Z/action/replication_record"}},"created_at":"2026-07-05T12:07:11.699034+00:00","updated_at":"2026-07-05T12:07:11.699034+00:00"}