{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:2RYTZS34WMZMQI5RTH7DGMX5UU","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2a8d2fcba6312f453f9406c66e80f6951f3c8790171ffa936b2134e3ff1ae125","cross_cats_sorted":["cs.CL","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-04-04T10:23:46Z","title_canon_sha256":"bcac663e8edd466d1911229db2b77e3a43cbca804b9ff5130cce9540ded13485"},"schema_version":"1.0","source":{"id":"2604.03675","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.03675","created_at":"2026-05-26T01:03:29Z"},{"alias_kind":"arxiv_version","alias_value":"2604.03675v3","created_at":"2026-05-26T01:03:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.03675","created_at":"2026-05-26T01:03:29Z"},{"alias_kind":"pith_short_12","alias_value":"2RYTZS34WMZM","created_at":"2026-05-26T01:03:29Z"},{"alias_kind":"pith_short_16","alias_value":"2RYTZS34WMZMQI5R","created_at":"2026-05-26T01:03:29Z"},{"alias_kind":"pith_short_8","alias_value":"2RYTZS34","created_at":"2026-05-26T01:03:29Z"}],"graph_snapshots":[{"event_id":"sha256:aaab8492a80512a9534346d462a57bd6f5c299e08d8a02313ae74ebbbb49eaa3","target":"graph","created_at":"2026-05-26T01:03:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Experiments on five multi-hop QA benchmarks show that OASES consistently outperforms strong RL baselines, with further analyses confirming the benefits of outcome-aligned process rewards and search-evaluation co-training."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That co-training the evaluator on the evolving policy will produce reliable, non-stale process rewards without introducing new biases or instability that undermine the final outcome alignment."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"OASES co-trains search policies and evaluators to generate outcome-aligned process rewards, outperforming standard RL baselines on five multi-hop QA benchmarks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"OASES improves agentic search by co-training policies with outcome-aligned evaluators for better process rewards."}],"snapshot_sha256":"55f67e31d805cec2b98399c31057da4981cf0b50f6bb63a6d9bd6005e6302495"},"formal_canon":{"evidence_count":1,"snapshot_sha256":"08e6c632355f317b811f748568968728fb6ae2cef81d28a9178b43a5bb8f8421"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2604.03675/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Agentic search enables language models to solve knowledge-intensive tasks by adaptively acquiring external evidence over multiple steps. Reinforcement learning with verifiable rewards (RLVR) has emerged as a widely adopted training paradigm for search agents, yet outcome-only rewards are sparse and provide limited credit assignment for intermediate search actions. Existing process-reward methods therefore seek to densify supervision through proxy signals, external evaluators, or likelihood-based information gain. However, proxy rewards can deviate from the final outcome objective, while fixed ","authors_text":"Erhan Zhang, Jiaxin Mao, Wei Yang, Xiaochi Wei, Yan Gao, Yao Hu, Yiqun Chen, Yi Wu, Zechun Niu","cross_cats":["cs.CL","cs.IR"],"headline":"OASES improves agentic search by co-training policies with outcome-aligned evaluators for better process rewards.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-04-04T10:23:46Z","title":"OASES: Outcome-Aligned Search-Evaluation Co-Training for Agentic Search"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.03675","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-13T17:12:43.396904Z","id":"b6697de3-c9bc-4f17-89f1-3de54d8a77fb","model_set":{"reader":"grok-4.3"},"one_line_summary":"OASES co-trains search policies and evaluators to generate outcome-aligned process rewards, outperforming standard RL baselines on five multi-hop QA benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"OASES improves agentic search by co-training policies with outcome-aligned evaluators for better process rewards.","strongest_claim":"Experiments on five multi-hop QA benchmarks show that OASES consistently outperforms strong RL baselines, with further analyses confirming the benefits of outcome-aligned process rewards and search-evaluation co-training.","weakest_assumption":"That co-training the evaluator on the evolving policy will produce reliable, non-stale process rewards without introducing new biases or instability that undermine the final outcome alignment."}},"verdict_id":"b6697de3-c9bc-4f17-89f1-3de54d8a77fb"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ca2d2d365613e452d08cbc7ee65db6b8583fa85a7d0883176851b97e04aa4eec","target":"record","created_at":"2026-05-26T01:03:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2a8d2fcba6312f453f9406c66e80f6951f3c8790171ffa936b2134e3ff1ae125","cross_cats_sorted":["cs.CL","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-04-04T10:23:46Z","title_canon_sha256":"bcac663e8edd466d1911229db2b77e3a43cbca804b9ff5130cce9540ded13485"},"schema_version":"1.0","source":{"id":"2604.03675","kind":"arxiv","version":3}},"canonical_sha256":"d4713ccb7cb332c823b199fe3332fda529f5dae2b9ae4a1527deb400f9dd1402","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d4713ccb7cb332c823b199fe3332fda529f5dae2b9ae4a1527deb400f9dd1402","first_computed_at":"2026-05-26T01:03:29.356430Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T01:03:29.356430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"w1y25hn3uspC8pjlp8Xf2FRQx+d/MV5U+XfCfB5+m0VpoChlPoBRw7cxuIuFMZwNjA8O9v+CpYx92gVLJzKzDg==","signature_status":"signed_v1","signed_at":"2026-05-26T01:03:29.357482Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.03675","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ca2d2d365613e452d08cbc7ee65db6b8583fa85a7d0883176851b97e04aa4eec","sha256:aaab8492a80512a9534346d462a57bd6f5c299e08d8a02313ae74ebbbb49eaa3"],"state_sha256":"4f98d70bf16480a3a1c962274b05878e0aceb62a564093144a519155f1b53300"}