{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:6UH7EGJORPNFNZLAT3AT5OR5BL","short_pith_number":"pith:6UH7EGJO","canonical_record":{"source":{"id":"2406.16086","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-23T11:57:53Z","cross_cats_sorted":[],"title_canon_sha256":"15d6b37d853e92d1261d0cfaaa141bfd442d298748d1a24d0c09af03f4044878","abstract_canon_sha256":"f23e5b7a50d81602f95cbb670caf06446472d69bd7dd71b500171ee98358f6fe"},"schema_version":"1.0"},"canonical_sha256":"f50ff2192e8bda56e5609ec13eba3d0af8ca133ca792cbcd16f87103bf59874d","source":{"kind":"arxiv","id":"2406.16086","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"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:6UH7EGJORPNFNZLAT3AT5OR5BL","target":"record","payload":{"canonical_record":{"source":{"id":"2406.16086","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-23T11:57:53Z","cross_cats_sorted":[],"title_canon_sha256":"15d6b37d853e92d1261d0cfaaa141bfd442d298748d1a24d0c09af03f4044878","abstract_canon_sha256":"f23e5b7a50d81602f95cbb670caf06446472d69bd7dd71b500171ee98358f6fe"},"schema_version":"1.0"},"canonical_sha256":"f50ff2192e8bda56e5609ec13eba3d0af8ca133ca792cbcd16f87103bf59874d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:35:53.127156Z","signature_b64":"gtm9gVgWmDQqZypHQvBGKTGedpWQs7dQnICJ9fLreamT7DWa+Cxdwzm0pRVeuh3L/KtYSJwNrdXuUVDgMn/4Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f50ff2192e8bda56e5609ec13eba3d0af8ca133ca792cbcd16f87103bf59874d","last_reissued_at":"2026-07-05T08:35:53.126698Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:35:53.126698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.16086","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:35:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"prgVjQC65KGyaMQ05dsaClFmbuigaHzCF5TW47+0z2xzhC0r/lapKD6Um8sDwssjA+xD3JXUvLRPqpRNIMZ8Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:39:18.399650Z"},"content_sha256":"9c786932ee869c0828453efa89bc7782feb5cdfb88dbd5cfcd6bbadea0f4caa5","schema_version":"1.0","event_id":"sha256:9c786932ee869c0828453efa89bc7782feb5cdfb88dbd5cfcd6bbadea0f4caa5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:6UH7EGJORPNFNZLAT3AT5OR5BL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SEAM: A Stochastic Benchmark for Multi-Document Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Arie Cattan, Avi Caciularu, Gabriel Stanovsky, Gili Lior, Ori Shapira, Shahar Levy","submitted_at":"2024-06-23T11:57:53Z","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"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.16086","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/2406.16086/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":"2026-07-05T08:35:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LwVjCZdAK143kc6S9QVhAuW6QkBF7AypNVDb9pUoWIO4NZDo7GqL+isVfBK0JBt12E3YwzYYlroy+vszEKQqBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:39:18.400142Z"},"content_sha256":"be9a5948fa3fb7f650925f57f9e3412bb0cf67875e9b0bf8c71ad91aa803d445","schema_version":"1.0","event_id":"sha256:be9a5948fa3fb7f650925f57f9e3412bb0cf67875e9b0bf8c71ad91aa803d445"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/bundle.json","state_url":"https://pith.science/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/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":"2026-08-09T00:39:18Z","links":{"resolver":"https://pith.science/pith/6UH7EGJORPNFNZLAT3AT5OR5BL","bundle":"https://pith.science/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/bundle.json","state":"https://pith.science/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6UH7EGJORPNFNZLAT3AT5OR5BL/bundle.json"},"state":{"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"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IeI70BviUNlrC5ZCRQCk0y8R1i06haSjN7OhiX+Mu95BCkdSebOjT6gJlFpAQBHEUUwq35hecv2kEnC3V2HhCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T00:39:18.404664Z","bundle_sha256":"d4dd83352fc161d1a41b221df27211f6c228d84551d84b519ac7accadb2581af"}}