{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:XSMADUCB6KDQDJLU5HZ3ZZMRX7","short_pith_number":"pith:XSMADUCB","canonical_record":{"source":{"id":"2409.02076","kind":"arxiv","version":7},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T17:25:54Z","cross_cats_sorted":[],"title_canon_sha256":"81934ffb77eb51c10e9e7a5e0005ddd52fdcd317106ae6d64277bce3579a886f","abstract_canon_sha256":"5571a553f3bbf447c18b7b6984f410be5b6babf8bec804c34a07cf1d49ed9239"},"schema_version":"1.0"},"canonical_sha256":"bc9801d041f28701a574e9f3bce591bffa1a9802383c425e10c3040738f0651a","source":{"kind":"arxiv","id":"2409.02076","version":7},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.02076","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"arxiv_version","alias_value":"2409.02076v7","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.02076","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_12","alias_value":"XSMADUCB6KDQ","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_16","alias_value":"XSMADUCB6KDQDJLU","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_8","alias_value":"XSMADUCB","created_at":"2026-07-05T10:04:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:XSMADUCB6KDQDJLU5HZ3ZZMRX7","target":"record","payload":{"canonical_record":{"source":{"id":"2409.02076","kind":"arxiv","version":7},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T17:25:54Z","cross_cats_sorted":[],"title_canon_sha256":"81934ffb77eb51c10e9e7a5e0005ddd52fdcd317106ae6d64277bce3579a886f","abstract_canon_sha256":"5571a553f3bbf447c18b7b6984f410be5b6babf8bec804c34a07cf1d49ed9239"},"schema_version":"1.0"},"canonical_sha256":"bc9801d041f28701a574e9f3bce591bffa1a9802383c425e10c3040738f0651a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:04:17.348239Z","signature_b64":"6yr9TmyyY+ddpXxCeypKU1LvQGH4NAhhx/SrZzBJEnyjFMIcPFfQs5wittGcV+23yMEWFYAg5ke+pXUoxElkCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bc9801d041f28701a574e9f3bce591bffa1a9802383c425e10c3040738f0651a","last_reissued_at":"2026-07-05T10:04:17.347788Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:04:17.347788Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.02076","source_version":7,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:04:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OwbB6cVx2Ru+2U6I1KanG7ZoX6unvmurRNoDtnXM+TmMM/CzYSktdnsMduyizDSGy6wkXBiXlSjlVxcWU+DAAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T16:55:23.216009Z"},"content_sha256":"131efc49e442de08b2ffcd8bc8d4f33985d6c518aca959b5abf1338c2c7de757","schema_version":"1.0","event_id":"sha256:131efc49e442de08b2ffcd8bc8d4f33985d6c518aca959b5abf1338c2c7de757"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:XSMADUCB6KDQDJLU5HZ3ZZMRX7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LongGenBench: Benchmarking Long-Form Generation in Long Context LLMs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ming Shan Hee, Roy Ka-Wei Lee, Yuhao Wu, Zhiqing Hu","submitted_at":"2024-09-03T17:25:54Z","abstract_excerpt":"Current benchmarks like Needle-in-a-Haystack (NIAH), Ruler, and Needlebench focus on models' ability to understand long-context input sequences but fail to capture a critical dimension: the generation of high-quality long-form text. Applications such as design proposals, technical documentation, and creative writing rely on coherent, instruction-following outputs over extended sequences - a challenge that existing benchmarks do not adequately address. To fill this gap, we introduce LongGenBench, a novel benchmark designed to rigorously evaluate large language models' (LLMs) ability to generate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.02076","kind":"arxiv","version":7},"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/2409.02076/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":"2026-07-05T10:04:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KgeAjN7dDfJesa7KMMDBdMmi6kKr48hMIDtgQoQNNT61qaU48if92xRl70IxNWn/RNZ3E46lSH9pZT4rFwhzDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T16:55:23.216568Z"},"content_sha256":"59179d6411deeff2effdca3e99f80de6d6895184824398db0da3b32c5ad1ae24","schema_version":"1.0","event_id":"sha256:59179d6411deeff2effdca3e99f80de6d6895184824398db0da3b32c5ad1ae24"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/bundle.json","state_url":"https://pith.science/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/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":"2026-08-10T16:55:23Z","links":{"resolver":"https://pith.science/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7","bundle":"https://pith.science/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/bundle.json","state":"https://pith.science/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XSMADUCB6KDQDJLU5HZ3ZZMRX7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XSMADUCB6KDQDJLU5HZ3ZZMRX7","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":"5571a553f3bbf447c18b7b6984f410be5b6babf8bec804c34a07cf1d49ed9239","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T17:25:54Z","title_canon_sha256":"81934ffb77eb51c10e9e7a5e0005ddd52fdcd317106ae6d64277bce3579a886f"},"schema_version":"1.0","source":{"id":"2409.02076","kind":"arxiv","version":7}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.02076","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"arxiv_version","alias_value":"2409.02076v7","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.02076","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_12","alias_value":"XSMADUCB6KDQ","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_16","alias_value":"XSMADUCB6KDQDJLU","created_at":"2026-07-05T10:04:17Z"},{"alias_kind":"pith_short_8","alias_value":"XSMADUCB","created_at":"2026-07-05T10:04:17Z"}],"graph_snapshots":[{"event_id":"sha256:59179d6411deeff2effdca3e99f80de6d6895184824398db0da3b32c5ad1ae24","target":"graph","created_at":"2026-07-05T10:04:17Z","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/2409.02076/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Current benchmarks like Needle-in-a-Haystack (NIAH), Ruler, and Needlebench focus on models' ability to understand long-context input sequences but fail to capture a critical dimension: the generation of high-quality long-form text. Applications such as design proposals, technical documentation, and creative writing rely on coherent, instruction-following outputs over extended sequences - a challenge that existing benchmarks do not adequately address. To fill this gap, we introduce LongGenBench, a novel benchmark designed to rigorously evaluate large language models' (LLMs) ability to generate","authors_text":"Ming Shan Hee, Roy Ka-Wei Lee, Yuhao Wu, Zhiqing Hu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T17:25:54Z","title":"LongGenBench: Benchmarking Long-Form Generation in Long Context LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.02076","kind":"arxiv","version":7},"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":"sha256:131efc49e442de08b2ffcd8bc8d4f33985d6c518aca959b5abf1338c2c7de757","target":"record","created_at":"2026-07-05T10:04:17Z","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":"5571a553f3bbf447c18b7b6984f410be5b6babf8bec804c34a07cf1d49ed9239","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-03T17:25:54Z","title_canon_sha256":"81934ffb77eb51c10e9e7a5e0005ddd52fdcd317106ae6d64277bce3579a886f"},"schema_version":"1.0","source":{"id":"2409.02076","kind":"arxiv","version":7}},"canonical_sha256":"bc9801d041f28701a574e9f3bce591bffa1a9802383c425e10c3040738f0651a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bc9801d041f28701a574e9f3bce591bffa1a9802383c425e10c3040738f0651a","first_computed_at":"2026-07-05T10:04:17.347788Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:04:17.347788Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6yr9TmyyY+ddpXxCeypKU1LvQGH4NAhhx/SrZzBJEnyjFMIcPFfQs5wittGcV+23yMEWFYAg5ke+pXUoxElkCg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:04:17.348239Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.02076","source_kind":"arxiv","source_version":7}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:131efc49e442de08b2ffcd8bc8d4f33985d6c518aca959b5abf1338c2c7de757","sha256:59179d6411deeff2effdca3e99f80de6d6895184824398db0da3b32c5ad1ae24"],"state_sha256":"5ec006f1dd529df1ba221026e3419e8ba1c03d5708daa5426a8dc626eddc3437"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NCRWcoZBojvzI9fnqr+fcDgae+5D2avRT7oX6yhNLLSTgTfsSHkcUFBk3NCE6C423DiDYzTHqXUkAPRFv+1bCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T16:55:23.221710Z","bundle_sha256":"253eb73c993774ff6aaa361019c9ffc057608d186f7cb0878b961658366bee51"}}