{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EH5OV6S7NIGQNZBGGNKXBMYIN2","short_pith_number":"pith:EH5OV6S7","schema_version":"1.0","canonical_sha256":"21faeafa5f6a0d06e426335570b3086ea35dc40750c63d3cea77383f56fb0a61","source":{"kind":"arxiv","id":"2605.19140","version":1},"attestation_state":"computed","paper":{"title":"Learning to Hand Off: Provably Convergent Workflow Learning under Interface Constraints","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Dawei Zhou, Elynn Chen, Enpei Zhang, Jiayu Li, Yujun Yan","submitted_at":"2026-05-18T21:48:02Z","abstract_excerpt":"We study workflow learning in a setting where specialized agents hand off control through a shared artifact, each agent observes only a local function of that artifact and its own private state, and no centralized learner accesses joint trajectories -- the operating regime of multi-agent LLM pipelines that span organizational, vendor, or trust boundaries. We formalize this regime as an interface-constrained semi-Markov decision process (IC-SMDP), whose decision epochs occur at handoff times, and design IC-$Q$, an asynchronous decentralized $Q$-learning algorithm in which cross-agent coordinati"},"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":"2605.19140","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-18T21:48:02Z","cross_cats_sorted":[],"title_canon_sha256":"7ca5c74dbb90b1e6ee5ed6957ec216a527747e5f7c7c570293cd28ac6e3a60ca","abstract_canon_sha256":"209690d8301753ff4ca7e812163ec18998faf60900a134b980a91d9e76372f59"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:29.229281Z","signature_b64":"GjriO43bDZUJ43xlaxlh52+W71O0G8Y37d0N9DThBpjkJhxshyR3botNfC5n6+RvzelP2vOtmD8gwQxb+dIBAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"21faeafa5f6a0d06e426335570b3086ea35dc40750c63d3cea77383f56fb0a61","last_reissued_at":"2026-05-20T01:05:29.228529Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:29.228529Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning to Hand Off: Provably Convergent Workflow Learning under Interface Constraints","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Dawei Zhou, Elynn Chen, Enpei Zhang, Jiayu Li, Yujun Yan","submitted_at":"2026-05-18T21:48:02Z","abstract_excerpt":"We study workflow learning in a setting where specialized agents hand off control through a shared artifact, each agent observes only a local function of that artifact and its own private state, and no centralized learner accesses joint trajectories -- the operating regime of multi-agent LLM pipelines that span organizational, vendor, or trust boundaries. We formalize this regime as an interface-constrained semi-Markov decision process (IC-SMDP), whose decision epochs occur at handoff times, and design IC-$Q$, an asynchronous decentralized $Q$-learning algorithm in which cross-agent coordinati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19140","kind":"arxiv","version":1},"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/2605.19140/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":"2605.19140","created_at":"2026-05-20T01:05:29.228649+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19140v1","created_at":"2026-05-20T01:05:29.228649+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19140","created_at":"2026-05-20T01:05:29.228649+00:00"},{"alias_kind":"pith_short_12","alias_value":"EH5OV6S7NIGQ","created_at":"2026-05-20T01:05:29.228649+00:00"},{"alias_kind":"pith_short_16","alias_value":"EH5OV6S7NIGQNZBG","created_at":"2026-05-20T01:05:29.228649+00:00"},{"alias_kind":"pith_short_8","alias_value":"EH5OV6S7","created_at":"2026-05-20T01:05:29.228649+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/EH5OV6S7NIGQNZBGGNKXBMYIN2","json":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2.json","graph_json":"https://pith.science/api/pith-number/EH5OV6S7NIGQNZBGGNKXBMYIN2/graph.json","events_json":"https://pith.science/api/pith-number/EH5OV6S7NIGQNZBGGNKXBMYIN2/events.json","paper":"https://pith.science/paper/EH5OV6S7"},"agent_actions":{"view_html":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2","download_json":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2.json","view_paper":"https://pith.science/paper/EH5OV6S7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19140&json=true","fetch_graph":"https://pith.science/api/pith-number/EH5OV6S7NIGQNZBGGNKXBMYIN2/graph.json","fetch_events":"https://pith.science/api/pith-number/EH5OV6S7NIGQNZBGGNKXBMYIN2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2/action/storage_attestation","attest_author":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2/action/author_attestation","sign_citation":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2/action/citation_signature","submit_replication":"https://pith.science/pith/EH5OV6S7NIGQNZBGGNKXBMYIN2/action/replication_record"}},"created_at":"2026-05-20T01:05:29.228649+00:00","updated_at":"2026-05-20T01:05:29.228649+00:00"}