{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:Z6NVJOR4Y76GON2DX7QMOB7UZM","short_pith_number":"pith:Z6NVJOR4","canonical_record":{"source":{"id":"2601.09515","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-14T14:31:16Z","cross_cats_sorted":[],"title_canon_sha256":"41fbef6f6d657150454dd098ce2d378035046898cc46f07e68114d419266f5fe","abstract_canon_sha256":"840e0aebb07e83d7414d725a272bc240de0d7e678df27774d68cbb4ad933c5a0"},"schema_version":"1.0"},"canonical_sha256":"cf9b54ba3cc7fc673743bfe0c707f4cb2101a73aca9eee147e1378ba8e404814","source":{"kind":"arxiv","id":"2601.09515","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.09515","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"arxiv_version","alias_value":"2601.09515v3","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.09515","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_12","alias_value":"Z6NVJOR4Y76G","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_16","alias_value":"Z6NVJOR4Y76GON2D","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_8","alias_value":"Z6NVJOR4","created_at":"2026-07-28T01:22:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:Z6NVJOR4Y76GON2DX7QMOB7UZM","target":"record","payload":{"canonical_record":{"source":{"id":"2601.09515","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-14T14:31:16Z","cross_cats_sorted":[],"title_canon_sha256":"41fbef6f6d657150454dd098ce2d378035046898cc46f07e68114d419266f5fe","abstract_canon_sha256":"840e0aebb07e83d7414d725a272bc240de0d7e678df27774d68cbb4ad933c5a0"},"schema_version":"1.0"},"canonical_sha256":"cf9b54ba3cc7fc673743bfe0c707f4cb2101a73aca9eee147e1378ba8e404814","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T01:22:27.716056Z","signature_b64":"1vF2/vCtBrxK9wvKze8Clc/g0tbQ640Myf2KYh746S3eJpiQ/6JhtF3HUrEWBk/42lQWT7gG+Gin/z02fwHzCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cf9b54ba3cc7fc673743bfe0c707f4cb2101a73aca9eee147e1378ba8e404814","last_reissued_at":"2026-07-28T01:22:27.715083Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T01:22:27.715083Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2601.09515","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-28T01:22:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NsUxyWMFfHTkTfj1WwEA/jgx2yxaWj2tUQxqLZn/KSR+4kVvQcfC6JFqHkzhItqUv/kfzh/sfkq02MC2k0etBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T06:14:56.484335Z"},"content_sha256":"fe66e9d3df1959668768aa91a6565c8e8d039a5ea382dec85c210160479a955f","schema_version":"1.0","event_id":"sha256:fe66e9d3df1959668768aa91a6565c8e8d039a5ea382dec85c210160479a955f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:Z6NVJOR4Y76GON2DX7QMOB7UZM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SERM: Self-Evolving Relevance Model with Agent-Driven Learning from Massive Query Streams","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Canjia Li, Chenglong Wang, Huiyu Wang, Jingbo Zhu, Qiaozhi He, Tianhua Zhou, Tong Xiao, Weixiong Lin, Xiaojia Chang, Xingzhao Zhu, Yifu Huo, Yun Yang","submitted_at":"2026-01-14T14:31:16Z","abstract_excerpt":"Due to the dynamically evolving nature of real-world query streams, relevance models struggle to generalize to practical search scenarios. A sophisticated solution is self-evolution techniques. However, in large-scale industrial settings with massive query streams, this technique faces two challenges: (1) informative samples are often sparse and difficult to identify, and (2) pseudo-labels generated by the current model could be unreliable. To address these challenges, in this work, we propose a Self-Evolving Relevance Model approach (SERM), which comprises two complementary multi-agent module"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"SERM can achieve significant performance gains through iterative self-evolution, as validated by extensive offline multilingual evaluations and online testing in a large-scale industrial setting serving billions of user requests daily.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The multi-agent sample miner can reliably detect distributional shifts and identify truly informative samples, and the multi-agent annotator can produce sufficiently reliable pseudo-labels despite the acknowledged sparsity and unreliability of labels generated by the current model.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SERM deploys multi-agent sample mining and two-level label agreement to enable iterative self-evolution of relevance models on industrial query streams, yielding performance gains in offline and online tests.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"40d95a17ccff72e7a42a32895905f2bef810b5a6342150bebc9784327814764c"},"source":{"id":"2601.09515","kind":"arxiv","version":3},"verdict":{"id":"1757ab49-6d9c-4b37-bf3e-18f648bbe88e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T14:37:42.194928Z","strongest_claim":"SERM can achieve significant performance gains through iterative self-evolution, as validated by extensive offline multilingual evaluations and online testing in a large-scale industrial setting serving billions of user requests daily.","one_line_summary":"SERM deploys multi-agent sample mining and two-level label agreement to enable iterative self-evolution of relevance models on industrial query streams, yielding performance gains in offline and online tests.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The multi-agent sample miner can reliably detect distributional shifts and identify truly informative samples, and the multi-agent annotator can produce sufficiently reliable pseudo-labels despite the acknowledged sparsity and unreliability of labels generated by the current model.","pith_extraction_headline":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.09515/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":2,"snapshot_sha256":"9e0487cf101059dbd04152747d5bf67b3e13756fbe07014787ed2a4cbab1fc2c"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"1757ab49-6d9c-4b37-bf3e-18f648bbe88e"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-28T01:22:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZrsP6Y+z+a0uFjDbh6hnGSdbMHnZw5zn41L2XpMn74xFQn+303gKzwviz8ihdl0r50WbJMwTAXQNSF+dz2R5Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T06:14:56.485039Z"},"content_sha256":"a1e9ff8e2831a45535a5a0f071928adad6a3d10d70ad709c947bb1714af85016","schema_version":"1.0","event_id":"sha256:a1e9ff8e2831a45535a5a0f071928adad6a3d10d70ad709c947bb1714af85016"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/bundle.json","state_url":"https://pith.science/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-12T06:14:56Z","links":{"resolver":"https://pith.science/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM","bundle":"https://pith.science/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/bundle.json","state":"https://pith.science/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Z6NVJOR4Y76GON2DX7QMOB7UZM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:Z6NVJOR4Y76GON2DX7QMOB7UZM","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":"840e0aebb07e83d7414d725a272bc240de0d7e678df27774d68cbb4ad933c5a0","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-14T14:31:16Z","title_canon_sha256":"41fbef6f6d657150454dd098ce2d378035046898cc46f07e68114d419266f5fe"},"schema_version":"1.0","source":{"id":"2601.09515","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.09515","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"arxiv_version","alias_value":"2601.09515v3","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.09515","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_12","alias_value":"Z6NVJOR4Y76G","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_16","alias_value":"Z6NVJOR4Y76GON2D","created_at":"2026-07-28T01:22:27Z"},{"alias_kind":"pith_short_8","alias_value":"Z6NVJOR4","created_at":"2026-07-28T01:22:27Z"}],"graph_snapshots":[{"event_id":"sha256:a1e9ff8e2831a45535a5a0f071928adad6a3d10d70ad709c947bb1714af85016","target":"graph","created_at":"2026-07-28T01:22:27Z","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":"SERM can achieve significant performance gains through iterative self-evolution, as validated by extensive offline multilingual evaluations and online testing in a large-scale industrial setting serving billions of user requests daily."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The multi-agent sample miner can reliably detect distributional shifts and identify truly informative samples, and the multi-agent annotator can produce sufficiently reliable pseudo-labels despite the acknowledged sparsity and unreliability of labels generated by the current model."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"SERM deploys multi-agent sample mining and two-level label agreement to enable iterative self-evolution of relevance models on industrial query streams, yielding performance gains in offline and online tests."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance."}],"snapshot_sha256":"40d95a17ccff72e7a42a32895905f2bef810b5a6342150bebc9784327814764c"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9e0487cf101059dbd04152747d5bf67b3e13756fbe07014787ed2a4cbab1fc2c"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2601.09515/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Due to the dynamically evolving nature of real-world query streams, relevance models struggle to generalize to practical search scenarios. A sophisticated solution is self-evolution techniques. However, in large-scale industrial settings with massive query streams, this technique faces two challenges: (1) informative samples are often sparse and difficult to identify, and (2) pseudo-labels generated by the current model could be unreliable. To address these challenges, in this work, we propose a Self-Evolving Relevance Model approach (SERM), which comprises two complementary multi-agent module","authors_text":"Canjia Li, Chenglong Wang, Huiyu Wang, Jingbo Zhu, Qiaozhi He, Tianhua Zhou, Tong Xiao, Weixiong Lin, Xiaojia Chang, Xingzhao Zhu, Yifu Huo, Yun Yang","cross_cats":[],"headline":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-14T14:31:16Z","title":"SERM: Self-Evolving Relevance Model with Agent-Driven Learning from Massive Query Streams"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.09515","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-16T14:37:42.194928Z","id":"1757ab49-6d9c-4b37-bf3e-18f648bbe88e","model_set":{"reader":"grok-4.3"},"one_line_summary":"SERM deploys multi-agent sample mining and two-level label agreement to enable iterative self-evolution of relevance models on industrial query streams, yielding performance gains in offline and online tests.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A multi-agent self-evolving model mines informative samples and generates reliable pseudo-labels from massive query streams to improve search relevance.","strongest_claim":"SERM can achieve significant performance gains through iterative self-evolution, as validated by extensive offline multilingual evaluations and online testing in a large-scale industrial setting serving billions of user requests daily.","weakest_assumption":"The multi-agent sample miner can reliably detect distributional shifts and identify truly informative samples, and the multi-agent annotator can produce sufficiently reliable pseudo-labels despite the acknowledged sparsity and unreliability of labels generated by the current model."}},"verdict_id":"1757ab49-6d9c-4b37-bf3e-18f648bbe88e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:fe66e9d3df1959668768aa91a6565c8e8d039a5ea382dec85c210160479a955f","target":"record","created_at":"2026-07-28T01:22:27Z","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":"840e0aebb07e83d7414d725a272bc240de0d7e678df27774d68cbb4ad933c5a0","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-14T14:31:16Z","title_canon_sha256":"41fbef6f6d657150454dd098ce2d378035046898cc46f07e68114d419266f5fe"},"schema_version":"1.0","source":{"id":"2601.09515","kind":"arxiv","version":3}},"canonical_sha256":"cf9b54ba3cc7fc673743bfe0c707f4cb2101a73aca9eee147e1378ba8e404814","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cf9b54ba3cc7fc673743bfe0c707f4cb2101a73aca9eee147e1378ba8e404814","first_computed_at":"2026-07-28T01:22:27.715083Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-28T01:22:27.715083Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1vF2/vCtBrxK9wvKze8Clc/g0tbQ640Myf2KYh746S3eJpiQ/6JhtF3HUrEWBk/42lQWT7gG+Gin/z02fwHzCg==","signature_status":"signed_v1","signed_at":"2026-07-28T01:22:27.716056Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.09515","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fe66e9d3df1959668768aa91a6565c8e8d039a5ea382dec85c210160479a955f","sha256:a1e9ff8e2831a45535a5a0f071928adad6a3d10d70ad709c947bb1714af85016"],"state_sha256":"7b90c2ba74e9c815001dcaeb93cf622931a93ed644bfbb191c981a00dfb551e8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8s+J9F5XIAbxB+2O30HqvrYlNytUf8KxAxIjJGN3RbnubPYdFejun4EL2R4C8Y6HI0vYYOAJ/b6kpZGCmGqfCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-12T06:14:56.490342Z","bundle_sha256":"07ff9fb301d0a5f39aa88db91491d89fad2a598e56086e3ddb5ddaa8daeaefe4"}}