{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:N5SOPHI4255B3HHX3DWJZZTOFZ","short_pith_number":"pith:N5SOPHI4","schema_version":"1.0","canonical_sha256":"6f64e79d1cd77a1d9cf7d8ec9ce66e2e7ef7848eee44348c54dc48331b6ccf4e","source":{"kind":"arxiv","id":"2204.04292","version":3},"attestation_state":"computed","paper":{"title":"Evolving Pareto-Optimal Actor-Critic Algorithms for Generalizability and Stability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aaron Parisi, Aleksandra Faust, Esteban Real, Jie Tan, John D. Co-Reyes, Juan Jose Garau-Luis, Yingjie Miao","submitted_at":"2022-04-08T20:46:16Z","abstract_excerpt":"Generalizability and stability are two key objectives for operating reinforcement learning (RL) agents in the real world. Designing RL algorithms that optimize these objectives can be a costly and painstaking process. This paper presents MetaPG, an evolutionary method for automated design of actor-critic loss functions. MetaPG explicitly optimizes for generalizability and performance, and implicitly optimizes the stability of both metrics. We initialize our loss function population with Soft Actor-Critic (SAC) and perform multi-objective optimization using fitness metrics encoding single-task "},"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":"2204.04292","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-04-08T20:46:16Z","cross_cats_sorted":[],"title_canon_sha256":"d1c34602e90bcc2897f747b1c545b0c64147d4b216863927c213b2846f57a988","abstract_canon_sha256":"c673864fb05ecbbceccf6ff4a1efccc565db06241e573183e792c9ac58f6fe70"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:03:53.888680Z","signature_b64":"tVuKoCXSuxmu0bYrcnN1TU36J/UZqwMxJ7Yw6mmOwpg0ijWl/hQ2q/oo6AxC7HtfCmAhcTm27WjkuYfGw/pzDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6f64e79d1cd77a1d9cf7d8ec9ce66e2e7ef7848eee44348c54dc48331b6ccf4e","last_reissued_at":"2026-07-05T06:03:53.888157Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:03:53.888157Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evolving Pareto-Optimal Actor-Critic Algorithms for Generalizability and Stability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aaron Parisi, Aleksandra Faust, Esteban Real, Jie Tan, John D. Co-Reyes, Juan Jose Garau-Luis, Yingjie Miao","submitted_at":"2022-04-08T20:46:16Z","abstract_excerpt":"Generalizability and stability are two key objectives for operating reinforcement learning (RL) agents in the real world. Designing RL algorithms that optimize these objectives can be a costly and painstaking process. This paper presents MetaPG, an evolutionary method for automated design of actor-critic loss functions. MetaPG explicitly optimizes for generalizability and performance, and implicitly optimizes the stability of both metrics. We initialize our loss function population with Soft Actor-Critic (SAC) and perform multi-objective optimization using fitness metrics encoding single-task "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2204.04292","kind":"arxiv","version":3},"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/2204.04292/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":"2204.04292","created_at":"2026-07-05T06:03:53.888218+00:00"},{"alias_kind":"arxiv_version","alias_value":"2204.04292v3","created_at":"2026-07-05T06:03:53.888218+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2204.04292","created_at":"2026-07-05T06:03:53.888218+00:00"},{"alias_kind":"pith_short_12","alias_value":"N5SOPHI4255B","created_at":"2026-07-05T06:03:53.888218+00:00"},{"alias_kind":"pith_short_16","alias_value":"N5SOPHI4255B3HHX","created_at":"2026-07-05T06:03:53.888218+00:00"},{"alias_kind":"pith_short_8","alias_value":"N5SOPHI4","created_at":"2026-07-05T06:03:53.888218+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/N5SOPHI4255B3HHX3DWJZZTOFZ","json":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ.json","graph_json":"https://pith.science/api/pith-number/N5SOPHI4255B3HHX3DWJZZTOFZ/graph.json","events_json":"https://pith.science/api/pith-number/N5SOPHI4255B3HHX3DWJZZTOFZ/events.json","paper":"https://pith.science/paper/N5SOPHI4"},"agent_actions":{"view_html":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ","download_json":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ.json","view_paper":"https://pith.science/paper/N5SOPHI4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2204.04292&json=true","fetch_graph":"https://pith.science/api/pith-number/N5SOPHI4255B3HHX3DWJZZTOFZ/graph.json","fetch_events":"https://pith.science/api/pith-number/N5SOPHI4255B3HHX3DWJZZTOFZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ/action/storage_attestation","attest_author":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ/action/author_attestation","sign_citation":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ/action/citation_signature","submit_replication":"https://pith.science/pith/N5SOPHI4255B3HHX3DWJZZTOFZ/action/replication_record"}},"created_at":"2026-07-05T06:03:53.888218+00:00","updated_at":"2026-07-05T06:03:53.888218+00:00"}