{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5PAI2YYOOYEZ45DTPYYM7JDFRV","short_pith_number":"pith:5PAI2YYO","schema_version":"1.0","canonical_sha256":"ebc08d630e76099e74737e30cfa4658d59fcf748d19c415c9e02d6c1ac0fd3ec","source":{"kind":"arxiv","id":"2603.17621","version":2},"attestation_state":"computed","paper":{"title":"Complementary RL: Towards Efficient Experience-Driven Agent Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bo Zheng, Dilxat Muhtar, Jiamang Wang, Jiashun Liu, Ju Huang, Ling Pan, Shaopan Xiong, Siran Yang, Wei Gao, Weixun Wang, Wenbo Su","submitted_at":"2026-03-18T11:38:01Z","abstract_excerpt":"Reinforcement Learning (RL) has emerged as a powerful paradigm for training LLM-based agents, yet remains limited by low sample efficiency, stemming not only from sparse outcome feedback but also from the agent's inability to leverage prior experience across episodes. While augmenting agents with historical experience offers a promising remedy, existing approaches suffer from a critical weakness: the experience distilled from history is either stored statically or fail to coevolve with the improving actor, causing a progressive misalignment between the experience and the actor's evolving capab"},"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":"2603.17621","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-18T11:38:01Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"9b87a23b53be89d72d44b4a16d7362932bb3c0732ba67c6ff4af5249ecebff00","abstract_canon_sha256":"78d94a22ea689ac777a4458adbb44c8d3058e6931b2f446f5a2d62cfeb4df221"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:16:29.400024Z","signature_b64":"zyqz+YLOZbLvxnewUsQAvXu376AzMzdCj6wGnfdD2aEL/tgxnMvJ3AwhuM4Cf2O0pVvLLMLtQPS5nOf28bF2DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ebc08d630e76099e74737e30cfa4658d59fcf748d19c415c9e02d6c1ac0fd3ec","last_reissued_at":"2026-06-30T01:16:29.399393Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:16:29.399393Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Complementary RL: Towards Efficient Experience-Driven Agent Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bo Zheng, Dilxat Muhtar, Jiamang Wang, Jiashun Liu, Ju Huang, Ling Pan, Shaopan Xiong, Siran Yang, Wei Gao, Weixun Wang, Wenbo Su","submitted_at":"2026-03-18T11:38:01Z","abstract_excerpt":"Reinforcement Learning (RL) has emerged as a powerful paradigm for training LLM-based agents, yet remains limited by low sample efficiency, stemming not only from sparse outcome feedback but also from the agent's inability to leverage prior experience across episodes. While augmenting agents with historical experience offers a promising remedy, existing approaches suffer from a critical weakness: the experience distilled from history is either stored statically or fail to coevolve with the improving actor, causing a progressive misalignment between the experience and the actor's evolving capab"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.17621","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/2603.17621/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":"2603.17621","created_at":"2026-06-30T01:16:29.399462+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.17621v2","created_at":"2026-06-30T01:16:29.399462+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.17621","created_at":"2026-06-30T01:16:29.399462+00:00"},{"alias_kind":"pith_short_12","alias_value":"5PAI2YYOOYEZ","created_at":"2026-06-30T01:16:29.399462+00:00"},{"alias_kind":"pith_short_16","alias_value":"5PAI2YYOOYEZ45DT","created_at":"2026-06-30T01:16:29.399462+00:00"},{"alias_kind":"pith_short_8","alias_value":"5PAI2YYO","created_at":"2026-06-30T01:16:29.399462+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":5,"sample":[{"citing_arxiv_id":"2605.01725","citing_title":"Motion-Aware Caching for Efficient Autoregressive Video Generation","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10663","citing_title":"Evolving-RL: End-to-End Optimization of Experience-Driven Self-Evolving Capability within Agents","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":20,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV","json":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV.json","graph_json":"https://pith.science/api/pith-number/5PAI2YYOOYEZ45DTPYYM7JDFRV/graph.json","events_json":"https://pith.science/api/pith-number/5PAI2YYOOYEZ45DTPYYM7JDFRV/events.json","paper":"https://pith.science/paper/5PAI2YYO"},"agent_actions":{"view_html":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV","download_json":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV.json","view_paper":"https://pith.science/paper/5PAI2YYO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.17621&json=true","fetch_graph":"https://pith.science/api/pith-number/5PAI2YYOOYEZ45DTPYYM7JDFRV/graph.json","fetch_events":"https://pith.science/api/pith-number/5PAI2YYOOYEZ45DTPYYM7JDFRV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV/action/storage_attestation","attest_author":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV/action/author_attestation","sign_citation":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV/action/citation_signature","submit_replication":"https://pith.science/pith/5PAI2YYOOYEZ45DTPYYM7JDFRV/action/replication_record"}},"created_at":"2026-06-30T01:16:29.399462+00:00","updated_at":"2026-06-30T01:16:29.399462+00:00"}