{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:6UH7EGJORPNFNZLAT3AT5OR5BL","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":"f23e5b7a50d81602f95cbb670caf06446472d69bd7dd71b500171ee98358f6fe","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-23T11:57:53Z","title_canon_sha256":"15d6b37d853e92d1261d0cfaaa141bfd442d298748d1a24d0c09af03f4044878"},"schema_version":"1.0","source":{"id":"2406.16086","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.16086","created_at":"2026-07-05T08:35:53Z"},{"alias_kind":"arxiv_version","alias_value":"2406.16086v1","created_at":"2026-07-05T08:35:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.16086","created_at":"2026-07-05T08:35:53Z"},{"alias_kind":"pith_short_12","alias_value":"6UH7EGJORPNF","created_at":"2026-07-05T08:35:53Z"},{"alias_kind":"pith_short_16","alias_value":"6UH7EGJORPNFNZLA","created_at":"2026-07-05T08:35:53Z"},{"alias_kind":"pith_short_8","alias_value":"6UH7EGJO","created_at":"2026-07-05T08:35:53Z"}],"graph_snapshots":[{"event_id":"sha256:be9a5948fa3fb7f650925f57f9e3412bb0cf67875e9b0bf8c71ad91aa803d445","target":"graph","created_at":"2026-07-05T08:35:53Z","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/2406.16086/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Various tasks, such as summarization, multi-hop question answering, or coreference resolution, are naturally phrased over collections of real-world documents. Such tasks present a unique set of challenges, revolving around the lack of coherent narrative structure across documents, which often leads to contradiction, omission, or repetition of information. Despite their real-world application and challenging properties, there is currently no benchmark which specifically measures the abilities of large language models (LLMs) on multi-document tasks. To bridge this gap, we present SEAM (a Stochas","authors_text":"Arie Cattan, Avi Caciularu, Gabriel Stanovsky, Gili Lior, Ori Shapira, Shahar Levy","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-23T11:57:53Z","title":"SEAM: A Stochastic Benchmark for Multi-Document Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.16086","kind":"arxiv","version":1},"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":"sha256:9c786932ee869c0828453efa89bc7782feb5cdfb88dbd5cfcd6bbadea0f4caa5","target":"record","created_at":"2026-07-05T08:35:53Z","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":"f23e5b7a50d81602f95cbb670caf06446472d69bd7dd71b500171ee98358f6fe","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-23T11:57:53Z","title_canon_sha256":"15d6b37d853e92d1261d0cfaaa141bfd442d298748d1a24d0c09af03f4044878"},"schema_version":"1.0","source":{"id":"2406.16086","kind":"arxiv","version":1}},"canonical_sha256":"f50ff2192e8bda56e5609ec13eba3d0af8ca133ca792cbcd16f87103bf59874d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f50ff2192e8bda56e5609ec13eba3d0af8ca133ca792cbcd16f87103bf59874d","first_computed_at":"2026-07-05T08:35:53.126698Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:35:53.126698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gtm9gVgWmDQqZypHQvBGKTGedpWQs7dQnICJ9fLreamT7DWa+Cxdwzm0pRVeuh3L/KtYSJwNrdXuUVDgMn/4Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:35:53.127156Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.16086","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9c786932ee869c0828453efa89bc7782feb5cdfb88dbd5cfcd6bbadea0f4caa5","sha256:be9a5948fa3fb7f650925f57f9e3412bb0cf67875e9b0bf8c71ad91aa803d445"],"state_sha256":"d6dff9ace4ec22ee20ea58f506e705234fd6ba36f9e08903f2f5c0119a914e6f"}