{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AON34PKGWAHTPVGCJGLKPEBR7T","short_pith_number":"pith:AON34PKG","schema_version":"1.0","canonical_sha256":"039bbe3d46b00f37d4c24996a79031fcf73614608fa6353ccb3ac59f81b449dc","source":{"kind":"arxiv","id":"2608.00030","version":1},"attestation_state":"computed","paper":{"title":"SLMs as Multi-Agent Routers: A Progressive SFT and Reinforcement Learning Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alexander Ng, Amir Kayhani, Gayathri V Kondapalli, Harvey Yorke, Hirsh Pithadia, Rahul Monish","submitted_at":"2026-07-15T11:11:39Z","abstract_excerpt":"Specialised retrieval agents typically surface higher quality results than general-purpose search, but selecting the optimal agent for a given query remains an open problem. Current approaches route queries based on inferred topic or intent, however intent-based selection is fundamentally limited: it does not incorporate signal from retrieved content, and cannot detect when a topically aligned agent produces low-relevance results. We address this by training a small language model via supervised fine-tuning followed by reinforcement learning to jointly perform agent selection and structured pa"},"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":"2608.00030","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-15T11:11:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e1688148281f7511ff36883196b3324ddd2fbdf35524c58b12cf36588cbebffa","abstract_canon_sha256":"952367391fd7e4671a0f2edd8b1f6f9ce45e8377ec48e0250c4fd9213e1eaaa2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T00:31:48.494839Z","signature_b64":"oxl6l4bZDlxTddndk+AY3CganluIDmEw1w5TQSXXKnvltsaHCyWi0fCK8mjdfWHPJqV1fYZUCLB5RbKdppTjDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"039bbe3d46b00f37d4c24996a79031fcf73614608fa6353ccb3ac59f81b449dc","last_reissued_at":"2026-08-04T00:31:48.493446Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T00:31:48.493446Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SLMs as Multi-Agent Routers: A Progressive SFT and Reinforcement Learning Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alexander Ng, Amir Kayhani, Gayathri V Kondapalli, Harvey Yorke, Hirsh Pithadia, Rahul Monish","submitted_at":"2026-07-15T11:11:39Z","abstract_excerpt":"Specialised retrieval agents typically surface higher quality results than general-purpose search, but selecting the optimal agent for a given query remains an open problem. Current approaches route queries based on inferred topic or intent, however intent-based selection is fundamentally limited: it does not incorporate signal from retrieved content, and cannot detect when a topically aligned agent produces low-relevance results. We address this by training a small language model via supervised fine-tuning followed by reinforcement learning to jointly perform agent selection and structured pa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.00030","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/2608.00030/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":"2608.00030","created_at":"2026-08-04T00:31:48.494456+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.00030v1","created_at":"2026-08-04T00:31:48.494456+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.00030","created_at":"2026-08-04T00:31:48.494456+00:00"},{"alias_kind":"pith_short_12","alias_value":"AON34PKGWAHT","created_at":"2026-08-04T00:31:48.494456+00:00"},{"alias_kind":"pith_short_16","alias_value":"AON34PKGWAHTPVGC","created_at":"2026-08-04T00:31:48.494456+00:00"},{"alias_kind":"pith_short_8","alias_value":"AON34PKG","created_at":"2026-08-04T00:31:48.494456+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/AON34PKGWAHTPVGCJGLKPEBR7T","json":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T.json","graph_json":"https://pith.science/api/pith-number/AON34PKGWAHTPVGCJGLKPEBR7T/graph.json","events_json":"https://pith.science/api/pith-number/AON34PKGWAHTPVGCJGLKPEBR7T/events.json","paper":"https://pith.science/paper/AON34PKG"},"agent_actions":{"view_html":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T","download_json":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T.json","view_paper":"https://pith.science/paper/AON34PKG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.00030&json=true","fetch_graph":"https://pith.science/api/pith-number/AON34PKGWAHTPVGCJGLKPEBR7T/graph.json","fetch_events":"https://pith.science/api/pith-number/AON34PKGWAHTPVGCJGLKPEBR7T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T/action/storage_attestation","attest_author":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T/action/author_attestation","sign_citation":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T/action/citation_signature","submit_replication":"https://pith.science/pith/AON34PKGWAHTPVGCJGLKPEBR7T/action/replication_record"}},"created_at":"2026-08-04T00:31:48.494456+00:00","updated_at":"2026-08-04T00:31:48.494456+00:00"}