{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZLMAXRMF7MC7NWCPZAF46UNCMK","short_pith_number":"pith:ZLMAXRMF","schema_version":"1.0","canonical_sha256":"cad80bc585fb05f6d84fc80bcf51a262b0ba49e4afa3398a2f1b1b08b96070d7","source":{"kind":"arxiv","id":"2508.01049","version":2},"attestation_state":"computed","paper":{"title":"Centralized Adaptive Sampling for Reliable Co-Training of Independent Multi-Agent Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Josiah P. Hanna, Nicholas E. Corrado","submitted_at":"2025-08-01T20:07:25Z","abstract_excerpt":"Independent on-policy policy gradient algorithms are widely used for multi-agent reinforcement learning (MARL) in cooperative and no-conflict games, but they are known to converge sub-optimally when each agent's individual policy gradient points away from an optimal joint equilibrium. Going beyond prior work, we observe that sub-optimal convergence can still arise even when the expected individual policy gradients of each agent point toward the optimal joint solution. After collecting a finite set of trajectories, stochasticity in independent action sampling can cause the joint data distributi"},"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":"2508.01049","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-01T20:07:25Z","cross_cats_sorted":[],"title_canon_sha256":"d160c0cbf665b54b7808ca28da06d3de60ea613d76d38d13c5e7d97ce500f235","abstract_canon_sha256":"bdaaf980e6e5efe9aba0d37a7c6f6e69f471a40d9ca4ba60d186675b6db1f6a6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:09:34.945516Z","signature_b64":"MUW7qvC2mi+hDIn8FW7wKB5HBCKi548KvdyR9XjfV1zrpXM1RmneC3WEvMd997SAZHqB7kIwmXYwTXTL0GLtCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cad80bc585fb05f6d84fc80bcf51a262b0ba49e4afa3398a2f1b1b08b96070d7","last_reissued_at":"2026-05-18T03:09:34.944961Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:09:34.944961Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Centralized Adaptive Sampling for Reliable Co-Training of Independent Multi-Agent Policies","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Josiah P. Hanna, Nicholas E. Corrado","submitted_at":"2025-08-01T20:07:25Z","abstract_excerpt":"Independent on-policy policy gradient algorithms are widely used for multi-agent reinforcement learning (MARL) in cooperative and no-conflict games, but they are known to converge sub-optimally when each agent's individual policy gradient points away from an optimal joint equilibrium. Going beyond prior work, we observe that sub-optimal convergence can still arise even when the expected individual policy gradients of each agent point toward the optimal joint solution. After collecting a finite set of trajectories, stochasticity in independent action sampling can cause the joint data distributi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.01049","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":""},"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":"2508.01049","created_at":"2026-05-18T03:09:34.945074+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.01049v2","created_at":"2026-05-18T03:09:34.945074+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.01049","created_at":"2026-05-18T03:09:34.945074+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZLMAXRMF7MC7","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZLMAXRMF7MC7NWCP","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZLMAXRMF","created_at":"2026-05-18T12:33:37.589309+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/ZLMAXRMF7MC7NWCPZAF46UNCMK","json":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK.json","graph_json":"https://pith.science/api/pith-number/ZLMAXRMF7MC7NWCPZAF46UNCMK/graph.json","events_json":"https://pith.science/api/pith-number/ZLMAXRMF7MC7NWCPZAF46UNCMK/events.json","paper":"https://pith.science/paper/ZLMAXRMF"},"agent_actions":{"view_html":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK","download_json":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK.json","view_paper":"https://pith.science/paper/ZLMAXRMF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.01049&json=true","fetch_graph":"https://pith.science/api/pith-number/ZLMAXRMF7MC7NWCPZAF46UNCMK/graph.json","fetch_events":"https://pith.science/api/pith-number/ZLMAXRMF7MC7NWCPZAF46UNCMK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK/action/storage_attestation","attest_author":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK/action/author_attestation","sign_citation":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK/action/citation_signature","submit_replication":"https://pith.science/pith/ZLMAXRMF7MC7NWCPZAF46UNCMK/action/replication_record"}},"created_at":"2026-05-18T03:09:34.945074+00:00","updated_at":"2026-05-18T03:09:34.945074+00:00"}