{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:2673ZVUXX2BRWYRLA7MCNVISLG","short_pith_number":"pith:2673ZVUX","canonical_record":{"source":{"id":"2312.09542","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T05:30:14Z","cross_cats_sorted":[],"title_canon_sha256":"02731e9a9f81117a2e92a84e3d970080c9a924544f21a7bc10de21ddb2f75db5","abstract_canon_sha256":"5180fb83e9160585bd8ba75326ab568fb6bbaf3ed6040a216d7986543da8f7a2"},"schema_version":"1.0"},"canonical_sha256":"d7bfbcd697be831b622b07d826d512599f870c6c455ce767cce23dd865e42f27","source":{"kind":"arxiv","id":"2312.09542","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.09542","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"arxiv_version","alias_value":"2312.09542v2","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.09542","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_12","alias_value":"2673ZVUXX2BR","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_16","alias_value":"2673ZVUXX2BRWYRL","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_8","alias_value":"2673ZVUX","created_at":"2026-07-05T08:36:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:2673ZVUXX2BRWYRLA7MCNVISLG","target":"record","payload":{"canonical_record":{"source":{"id":"2312.09542","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T05:30:14Z","cross_cats_sorted":[],"title_canon_sha256":"02731e9a9f81117a2e92a84e3d970080c9a924544f21a7bc10de21ddb2f75db5","abstract_canon_sha256":"5180fb83e9160585bd8ba75326ab568fb6bbaf3ed6040a216d7986543da8f7a2"},"schema_version":"1.0"},"canonical_sha256":"d7bfbcd697be831b622b07d826d512599f870c6c455ce767cce23dd865e42f27","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:44.963279Z","signature_b64":"Ybu9f1DjuOZkv9Eo0Zh8Q7Bp/PQ5OAy0yPdZenBqfIT/YxJIfRPZi8Jk2KJVRoJ8wBLTBxOKWbAnK8Pwyf+GAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7bfbcd697be831b622b07d826d512599f870c6c455ce767cce23dd865e42f27","last_reissued_at":"2026-07-05T08:36:44.962790Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:44.962790Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2312.09542","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":"2026-07-05T08:36:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1n3uyc0CdZOwWyAYHegDtSfhzWVo1V3H7EMik35XX5k5rwKLQENgK809AScrZIymfQw629c/afkFkdI8CmmsDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T08:29:03.676850Z"},"content_sha256":"4a71812ccecf29bfad31bc3ff633208e3b94c68fd09b90bc79b4df34d4a12cf6","schema_version":"1.0","event_id":"sha256:4a71812ccecf29bfad31bc3ff633208e3b94c68fd09b90bc79b4df34d4a12cf6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:2673ZVUXX2BRWYRLA7MCNVISLG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Marathon: A Race Through the Realm of Long Context with Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jiaxi Yang, Junhao Liu, Lei Zhang, Longze Chen, Min Yang, Run Luo, Yunshui Li, Ziqiang Liu","submitted_at":"2023-12-15T05:30:14Z","abstract_excerpt":"With the advancement of large language models (LLMs) and the expansion of their context windows, existing long-context benchmarks fall short in effectively evaluating the models' comprehension and reasoning abilities in extended texts. Moreover, conventional benchmarks relying on F1 metrics often inaccurately score responses: they may undervalue correct answers that differ from the reference responses and overvalue incorrect ones that resemble the reference texts. In response to these limitations, we introduce Marathon, a novel evaluation benchmark that adopts a multiple-choice question format"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.09542","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/2312.09542/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":"2026-07-05T08:36:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LKNd8v6ekxLL63H8iv3uX2lWsXMcllrCGEMl5lHuGt0X+T6yoeQCPb6GihYLbLyMYAoqaco1ReKJ43Sp27BlAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T08:29:03.677588Z"},"content_sha256":"65fec7390c31f82b2e1ffd78fbb21047e41243fcd03210fe09dfc038f7c48dd8","schema_version":"1.0","event_id":"sha256:65fec7390c31f82b2e1ffd78fbb21047e41243fcd03210fe09dfc038f7c48dd8"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2673ZVUXX2BRWYRLA7MCNVISLG/bundle.json","state_url":"https://pith.science/pith/2673ZVUXX2BRWYRLA7MCNVISLG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2673ZVUXX2BRWYRLA7MCNVISLG/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":"2026-08-21T08:29:03Z","links":{"resolver":"https://pith.science/pith/2673ZVUXX2BRWYRLA7MCNVISLG","bundle":"https://pith.science/pith/2673ZVUXX2BRWYRLA7MCNVISLG/bundle.json","state":"https://pith.science/pith/2673ZVUXX2BRWYRLA7MCNVISLG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2673ZVUXX2BRWYRLA7MCNVISLG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:2673ZVUXX2BRWYRLA7MCNVISLG","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":"5180fb83e9160585bd8ba75326ab568fb6bbaf3ed6040a216d7986543da8f7a2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T05:30:14Z","title_canon_sha256":"02731e9a9f81117a2e92a84e3d970080c9a924544f21a7bc10de21ddb2f75db5"},"schema_version":"1.0","source":{"id":"2312.09542","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.09542","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"arxiv_version","alias_value":"2312.09542v2","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.09542","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_12","alias_value":"2673ZVUXX2BR","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_16","alias_value":"2673ZVUXX2BRWYRL","created_at":"2026-07-05T08:36:44Z"},{"alias_kind":"pith_short_8","alias_value":"2673ZVUX","created_at":"2026-07-05T08:36:44Z"}],"graph_snapshots":[{"event_id":"sha256:65fec7390c31f82b2e1ffd78fbb21047e41243fcd03210fe09dfc038f7c48dd8","target":"graph","created_at":"2026-07-05T08:36:44Z","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/2312.09542/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"With the advancement of large language models (LLMs) and the expansion of their context windows, existing long-context benchmarks fall short in effectively evaluating the models' comprehension and reasoning abilities in extended texts. Moreover, conventional benchmarks relying on F1 metrics often inaccurately score responses: they may undervalue correct answers that differ from the reference responses and overvalue incorrect ones that resemble the reference texts. In response to these limitations, we introduce Marathon, a novel evaluation benchmark that adopts a multiple-choice question format","authors_text":"Jiaxi Yang, Junhao Liu, Lei Zhang, Longze Chen, Min Yang, Run Luo, Yunshui Li, Ziqiang Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T05:30:14Z","title":"Marathon: A Race Through the Realm of Long Context with Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.09542","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":"sha256:4a71812ccecf29bfad31bc3ff633208e3b94c68fd09b90bc79b4df34d4a12cf6","target":"record","created_at":"2026-07-05T08:36:44Z","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":"5180fb83e9160585bd8ba75326ab568fb6bbaf3ed6040a216d7986543da8f7a2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T05:30:14Z","title_canon_sha256":"02731e9a9f81117a2e92a84e3d970080c9a924544f21a7bc10de21ddb2f75db5"},"schema_version":"1.0","source":{"id":"2312.09542","kind":"arxiv","version":2}},"canonical_sha256":"d7bfbcd697be831b622b07d826d512599f870c6c455ce767cce23dd865e42f27","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d7bfbcd697be831b622b07d826d512599f870c6c455ce767cce23dd865e42f27","first_computed_at":"2026-07-05T08:36:44.962790Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:36:44.962790Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ybu9f1DjuOZkv9Eo0Zh8Q7Bp/PQ5OAy0yPdZenBqfIT/YxJIfRPZi8Jk2KJVRoJ8wBLTBxOKWbAnK8Pwyf+GAw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:36:44.963279Z","signed_message":"canonical_sha256_bytes"},"source_id":"2312.09542","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4a71812ccecf29bfad31bc3ff633208e3b94c68fd09b90bc79b4df34d4a12cf6","sha256:65fec7390c31f82b2e1ffd78fbb21047e41243fcd03210fe09dfc038f7c48dd8"],"state_sha256":"a6daaf5ec4fec9278338c2ac0edaefa16d54f2ee33d833fd4d4ae54c7d320fae"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fA9pE9tIHZ/AzEWBDBzSruen5qS/FYx92nw2eoOtke6ce4kYqT7L8v7nog/I7owNIFzbfdk6TENXqb1Enlu7DA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-21T08:29:03.682849Z","bundle_sha256":"c1341ea163cc8b9f879d057312c29a25961193d4cbd5d17f706a48d96dd3f58c"}}