{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:LLQV56A4ELKCKYUKA2K3YGPIXD","short_pith_number":"pith:LLQV56A4","canonical_record":{"source":{"id":"2506.21506","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-26T17:32:50Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"45767dd7349f1535532149b2b0c624d5d2ef8851466cbf27760d801fd52b8a1e","abstract_canon_sha256":"868783083ade9dc53daa40347ca1f1227acddbc09db5a065c171b6e77ddf770e"},"schema_version":"1.0"},"canonical_sha256":"5ae15ef81c22d425628a0695bc19e8b8ec2cba324ec244f3acca1dc2148447e3","source":{"kind":"arxiv","id":"2506.21506","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.21506","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"arxiv_version","alias_value":"2506.21506v2","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.21506","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_12","alias_value":"LLQV56A4ELKC","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_16","alias_value":"LLQV56A4ELKCKYUK","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_8","alias_value":"LLQV56A4","created_at":"2026-07-05T11:31:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:LLQV56A4ELKCKYUKA2K3YGPIXD","target":"record","payload":{"canonical_record":{"source":{"id":"2506.21506","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-26T17:32:50Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"45767dd7349f1535532149b2b0c624d5d2ef8851466cbf27760d801fd52b8a1e","abstract_canon_sha256":"868783083ade9dc53daa40347ca1f1227acddbc09db5a065c171b6e77ddf770e"},"schema_version":"1.0"},"canonical_sha256":"5ae15ef81c22d425628a0695bc19e8b8ec2cba324ec244f3acca1dc2148447e3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:31:28.877772Z","signature_b64":"3pQySJr76oGqdUeYlssea8fWotnGWArMRzWhhl94lqTlKFKHnJx+Th1xG5TjD1S/b3xSPmE6PgejTsajXgKBCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ae15ef81c22d425628a0695bc19e8b8ec2cba324ec244f3acca1dc2148447e3","last_reissued_at":"2026-07-05T11:31:28.877250Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:31:28.877250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.21506","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:31:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QXProM5EdcxrZg1UvBy897RnQIDePLiYdoD51NcxTr81g+d6fz1YBEo29PgMzHLfC/ZkzRoek+wQ/vt1wqbwBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T11:43:17.423076Z"},"content_sha256":"bc30feb81b8303000613b436f0ddb2605d3fbb187f68f5f14b7b4af4458d39b1","schema_version":"1.0","event_id":"sha256:bc30feb81b8303000613b436f0ddb2605d3fbb187f68f5f14b7b4af4458d39b1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:LLQV56A4ELKCKYUKA2K3YGPIXD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Andrei Kopanev, Bernal Jim\\'enez Guti\\'errez, Botao Yu, Boyuan Zheng, Boyu Gou, Chan Hee Song, Hanane Nour Moussa, Huan Sun, Jiaman Wu, Jian Xie, Kai Zhang, Michael Lin, Morteza Ziyadi, Shijie Chen, Tianci Xue, Tianshu Zhang, Viktor Rozgic, Weijian Qi, Yifei Li, Yiheng Shu, Yu Gu, Yu Su, Yuting Ning, Zanming Huang, Zeyi Liao, Zhaowei Cai","submitted_at":"2025-06-26T17:32:50Z","abstract_excerpt":"Agentic search such as Deep Research systems-where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers-represents a major shift in how users interact with web-scale information. While promising greater efficiency and cognitive offloading, the growing complexity and open-endedness of agentic search have outpaced existing evaluation benchmarks and methodologies, which largely assume short search horizons and static answers. In this paper, we introduce Mind2Web 2, a benchmark of 130 realistic, high-quality, and long-horizon tasks that requi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.21506","kind":"arxiv","version":2},"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/2506.21506/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:31:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZTJjcD2qqMLXRHaAZYriJ9qysz2A8tDIiVU6H+iCn+iavWDERQK4vzT+g+CnCwW4s3S2Whn2uAHppBKy/pUBBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T11:43:17.424058Z"},"content_sha256":"6c618a7bae738669d85baff3d3c8d34b245ad580bbb041fe058cdcacca97a0d0","schema_version":"1.0","event_id":"sha256:6c618a7bae738669d85baff3d3c8d34b245ad580bbb041fe058cdcacca97a0d0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/bundle.json","state_url":"https://pith.science/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/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":"2026-08-05T11:43:17Z","links":{"resolver":"https://pith.science/pith/LLQV56A4ELKCKYUKA2K3YGPIXD","bundle":"https://pith.science/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/bundle.json","state":"https://pith.science/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LLQV56A4ELKCKYUKA2K3YGPIXD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:LLQV56A4ELKCKYUKA2K3YGPIXD","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":"868783083ade9dc53daa40347ca1f1227acddbc09db5a065c171b6e77ddf770e","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-26T17:32:50Z","title_canon_sha256":"45767dd7349f1535532149b2b0c624d5d2ef8851466cbf27760d801fd52b8a1e"},"schema_version":"1.0","source":{"id":"2506.21506","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.21506","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"arxiv_version","alias_value":"2506.21506v2","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.21506","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_12","alias_value":"LLQV56A4ELKC","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_16","alias_value":"LLQV56A4ELKCKYUK","created_at":"2026-07-05T11:31:28Z"},{"alias_kind":"pith_short_8","alias_value":"LLQV56A4","created_at":"2026-07-05T11:31:28Z"}],"graph_snapshots":[{"event_id":"sha256:6c618a7bae738669d85baff3d3c8d34b245ad580bbb041fe058cdcacca97a0d0","target":"graph","created_at":"2026-07-05T11:31:28Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2506.21506/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Agentic search such as Deep Research systems-where agents autonomously browse the web, synthesize information, and return comprehensive citation-backed answers-represents a major shift in how users interact with web-scale information. While promising greater efficiency and cognitive offloading, the growing complexity and open-endedness of agentic search have outpaced existing evaluation benchmarks and methodologies, which largely assume short search horizons and static answers. In this paper, we introduce Mind2Web 2, a benchmark of 130 realistic, high-quality, and long-horizon tasks that requi","authors_text":"Andrei Kopanev, Bernal Jim\\'enez Guti\\'errez, Botao Yu, Boyuan Zheng, Boyu Gou, Chan Hee Song, Hanane Nour Moussa, Huan Sun, Jiaman Wu, Jian Xie, Kai Zhang, Michael Lin, Morteza Ziyadi, Shijie Chen, Tianci Xue, Tianshu Zhang, Viktor Rozgic, Weijian Qi, Yifei Li, Yiheng Shu, Yu Gu, Yu Su, Yuting Ning, Zanming Huang, Zeyi Liao, Zhaowei Cai","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-26T17:32:50Z","title":"Mind2Web 2: Evaluating Agentic Search with Agent-as-a-Judge"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.21506","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:bc30feb81b8303000613b436f0ddb2605d3fbb187f68f5f14b7b4af4458d39b1","target":"record","created_at":"2026-07-05T11:31:28Z","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":"868783083ade9dc53daa40347ca1f1227acddbc09db5a065c171b6e77ddf770e","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-26T17:32:50Z","title_canon_sha256":"45767dd7349f1535532149b2b0c624d5d2ef8851466cbf27760d801fd52b8a1e"},"schema_version":"1.0","source":{"id":"2506.21506","kind":"arxiv","version":2}},"canonical_sha256":"5ae15ef81c22d425628a0695bc19e8b8ec2cba324ec244f3acca1dc2148447e3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5ae15ef81c22d425628a0695bc19e8b8ec2cba324ec244f3acca1dc2148447e3","first_computed_at":"2026-07-05T11:31:28.877250Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:31:28.877250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"3pQySJr76oGqdUeYlssea8fWotnGWArMRzWhhl94lqTlKFKHnJx+Th1xG5TjD1S/b3xSPmE6PgejTsajXgKBCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:31:28.877772Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.21506","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bc30feb81b8303000613b436f0ddb2605d3fbb187f68f5f14b7b4af4458d39b1","sha256:6c618a7bae738669d85baff3d3c8d34b245ad580bbb041fe058cdcacca97a0d0"],"state_sha256":"42d1058756f4b1d8b05c08424295eb8cb39af5a879a2ed3b3298cdcc11c205d9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"J2q4PD+d29lUqDW9cRiTDOz66vRk/4UvqQF+B/m5W/dN0+iC9JBnCu5DS+7IQU18t0S9aMQWnH0I0e6s6aMMAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T11:43:17.435229Z","bundle_sha256":"ec2c69674fe7dd809bbd9d365b87965ea90ba32048eb6d3fa54df8399ca46b33"}}