{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BVQGL67QBSNCNCH7BC3ME5B74P","short_pith_number":"pith:BVQGL67Q","schema_version":"1.0","canonical_sha256":"0d6065fbf00c9a2688ff08b6c2743fe3c9af2d6ec2f973fcf9372591293719ce","source":{"kind":"arxiv","id":"2504.16129","version":5},"attestation_state":"computed","paper":{"title":"MARFT: Multi-Agent Reinforcement Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.RO"],"primary_cat":"cs.MA","authors_text":"Jun Wang, Junwei Liao, Muning Wen, Weinan Zhang","submitted_at":"2025-04-21T07:03:54Z","abstract_excerpt":"Large Language Model (LLM)-based Multi-Agent Systems (LaMAS) have demonstrated strong capabilities on complex agentic tasks requiring multifaceted reasoning and collaboration, from high-quality presentation generation to scientific research. Meanwhile, Reinforcement Learning (RL) is widely recognized for enhancing agent intelligence, but limited work has studied fine-tuning LaMAS with foundational RL techniques. Directly applying conventional Multi-Agent Reinforcement Learning (MARL) to LaMAS also introduces major challenges due to the unique mechanisms of LaMAS. To address these challenges, t"},"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":"2504.16129","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MA","submitted_at":"2025-04-21T07:03:54Z","cross_cats_sorted":["cs.AI","cs.LG","cs.RO"],"title_canon_sha256":"61fdee7f5c822a2f7cf8d5c84aca2b5db9bf3e79c8c70b20d35e6db90d1ec21e","abstract_canon_sha256":"336e0a9c7647277250c62c3de943c9635a57288917e24209a2b1b2589dfce52b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:31.278690Z","signature_b64":"Rcrl6bpbiPNcaB8S/KKz3v18Nt8p+RwDvm/S/aEPbZcxVTGldroETaFI2aDuDJWC6WzbHY9Gm4kJKf54gW9gCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0d6065fbf00c9a2688ff08b6c2743fe3c9af2d6ec2f973fcf9372591293719ce","last_reissued_at":"2026-06-02T01:03:31.278171Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:31.278171Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MARFT: Multi-Agent Reinforcement Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.RO"],"primary_cat":"cs.MA","authors_text":"Jun Wang, Junwei Liao, Muning Wen, Weinan Zhang","submitted_at":"2025-04-21T07:03:54Z","abstract_excerpt":"Large Language Model (LLM)-based Multi-Agent Systems (LaMAS) have demonstrated strong capabilities on complex agentic tasks requiring multifaceted reasoning and collaboration, from high-quality presentation generation to scientific research. Meanwhile, Reinforcement Learning (RL) is widely recognized for enhancing agent intelligence, but limited work has studied fine-tuning LaMAS with foundational RL techniques. Directly applying conventional Multi-Agent Reinforcement Learning (MARL) to LaMAS also introduces major challenges due to the unique mechanisms of LaMAS. To address these challenges, t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.16129","kind":"arxiv","version":5},"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/2504.16129/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":"2504.16129","created_at":"2026-06-02T01:03:31.278243+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.16129v5","created_at":"2026-06-02T01:03:31.278243+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.16129","created_at":"2026-06-02T01:03:31.278243+00:00"},{"alias_kind":"pith_short_12","alias_value":"BVQGL67QBSNC","created_at":"2026-06-02T01:03:31.278243+00:00"},{"alias_kind":"pith_short_16","alias_value":"BVQGL67QBSNCNCH7","created_at":"2026-06-02T01:03:31.278243+00:00"},{"alias_kind":"pith_short_8","alias_value":"BVQGL67Q","created_at":"2026-06-02T01:03:31.278243+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":12,"sample":[{"citing_arxiv_id":"2504.01990","citing_title":"Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems","ref_index":168,"is_internal_anchor":true},{"citing_arxiv_id":"2601.22297","citing_title":"Learning from Self-Debate: Preparing Reasoning Models for Multi-Agent Debate","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08401","citing_title":"AIPO: Learning to Reason from Active Interaction","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2601.21972","citing_title":"Learning Decentralized LLM Collaboration with Multi-Agent Actor Critic","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2508.07407","citing_title":"A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2603.12631","citing_title":"Joint Optimization of Multi-agent Memory System","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12943","citing_title":"Reinforced Collaboration in Multi-Agent Flow Networks","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12966","citing_title":"Position: Agentic AI System Is a Foreseeable Pathway to AGI","ref_index":41,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08401","citing_title":"AIPO: Learning to Reason from Active Interaction","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04811","citing_title":"Tree-based Credit Assignment for Multi-Agent Memory System","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13564","citing_title":"Memory in the Age of AI Agents","ref_index":67,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02801","citing_title":"Reinforcement Learning for LLM-based Multi-Agent Systems through Orchestration Traces","ref_index":32,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P","json":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P.json","graph_json":"https://pith.science/api/pith-number/BVQGL67QBSNCNCH7BC3ME5B74P/graph.json","events_json":"https://pith.science/api/pith-number/BVQGL67QBSNCNCH7BC3ME5B74P/events.json","paper":"https://pith.science/paper/BVQGL67Q"},"agent_actions":{"view_html":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P","download_json":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P.json","view_paper":"https://pith.science/paper/BVQGL67Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.16129&json=true","fetch_graph":"https://pith.science/api/pith-number/BVQGL67QBSNCNCH7BC3ME5B74P/graph.json","fetch_events":"https://pith.science/api/pith-number/BVQGL67QBSNCNCH7BC3ME5B74P/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P/action/storage_attestation","attest_author":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P/action/author_attestation","sign_citation":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P/action/citation_signature","submit_replication":"https://pith.science/pith/BVQGL67QBSNCNCH7BC3ME5B74P/action/replication_record"}},"created_at":"2026-06-02T01:03:31.278243+00:00","updated_at":"2026-06-02T01:03:31.278243+00:00"}