{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SGIXCAWFXAON2GDGCWOCOKMVMH","short_pith_number":"pith:SGIXCAWF","schema_version":"1.0","canonical_sha256":"91917102c5b81cdd1866159c27299561fdc3ffd674fdaf42e0281fc46d63f29e","source":{"kind":"arxiv","id":"2606.10709","version":1},"attestation_state":"computed","paper":{"title":"Effective Reinforcement Learning for Agentic Search by Recycling Zero-Variance Queries During Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Bruno Martins, Chenyan Xiong, Jo\\~ao Coelho, Jo\\~ao Magalh\\~aes","submitted_at":"2026-06-09T11:12:58Z","abstract_excerpt":"The use of GRPO-style algorithms has become the standard strategy for training LLM search agents under outcome-only rewards. With these algorithms, a query contributes to parameter updates only when its rollout group mixes successes and failures; all-correct (too-easy) and all-incorrect (too-hard) groups are zero-variance and waste rollout cost. Existing approaches treat zero-variance as a static property and either discard or pre-filter such groups. We hypothesize and empirically validate that queries flip between zero-variance and signal-bearing states as the policy evolves during training. "},"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":"2606.10709","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2026-06-09T11:12:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f57973817aff724c555f92e3e811ef87a3652c8554edd4b310de9455134b1cd5","abstract_canon_sha256":"5c2673c3022203c966059d64d2d72a4f0321088d908b7e17af32ab0a0b09e740"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:35.635893Z","signature_b64":"MmdlTaV9NERUxpg2j7DmKg2akzJD95fR4ph9kscaoCsfgNwGBfAIbCUwTqeFF8QVBEbB7eQdM1R5LuRRyxHgCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91917102c5b81cdd1866159c27299561fdc3ffd674fdaf42e0281fc46d63f29e","last_reissued_at":"2026-06-10T01:10:35.634984Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:35.634984Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Effective Reinforcement Learning for Agentic Search by Recycling Zero-Variance Queries During Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Bruno Martins, Chenyan Xiong, Jo\\~ao Coelho, Jo\\~ao Magalh\\~aes","submitted_at":"2026-06-09T11:12:58Z","abstract_excerpt":"The use of GRPO-style algorithms has become the standard strategy for training LLM search agents under outcome-only rewards. With these algorithms, a query contributes to parameter updates only when its rollout group mixes successes and failures; all-correct (too-easy) and all-incorrect (too-hard) groups are zero-variance and waste rollout cost. Existing approaches treat zero-variance as a static property and either discard or pre-filter such groups. We hypothesize and empirically validate that queries flip between zero-variance and signal-bearing states as the policy evolves during training. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10709","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/2606.10709/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":"2606.10709","created_at":"2026-06-10T01:10:35.635130+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10709v1","created_at":"2026-06-10T01:10:35.635130+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10709","created_at":"2026-06-10T01:10:35.635130+00:00"},{"alias_kind":"pith_short_12","alias_value":"SGIXCAWFXAON","created_at":"2026-06-10T01:10:35.635130+00:00"},{"alias_kind":"pith_short_16","alias_value":"SGIXCAWFXAON2GDG","created_at":"2026-06-10T01:10:35.635130+00:00"},{"alias_kind":"pith_short_8","alias_value":"SGIXCAWF","created_at":"2026-06-10T01:10:35.635130+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/SGIXCAWFXAON2GDGCWOCOKMVMH","json":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH.json","graph_json":"https://pith.science/api/pith-number/SGIXCAWFXAON2GDGCWOCOKMVMH/graph.json","events_json":"https://pith.science/api/pith-number/SGIXCAWFXAON2GDGCWOCOKMVMH/events.json","paper":"https://pith.science/paper/SGIXCAWF"},"agent_actions":{"view_html":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH","download_json":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH.json","view_paper":"https://pith.science/paper/SGIXCAWF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10709&json=true","fetch_graph":"https://pith.science/api/pith-number/SGIXCAWFXAON2GDGCWOCOKMVMH/graph.json","fetch_events":"https://pith.science/api/pith-number/SGIXCAWFXAON2GDGCWOCOKMVMH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH/action/storage_attestation","attest_author":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH/action/author_attestation","sign_citation":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH/action/citation_signature","submit_replication":"https://pith.science/pith/SGIXCAWFXAON2GDGCWOCOKMVMH/action/replication_record"}},"created_at":"2026-06-10T01:10:35.635130+00:00","updated_at":"2026-06-10T01:10:35.635130+00:00"}