{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:K23ULI54QW6YEBILIMWIU75266","short_pith_number":"pith:K23ULI54","canonical_record":{"source":{"id":"2412.13670","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-18T09:53:12Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4ebf8ff6c99a7cc4f356bc26440de1bff54163aa8c2b9622272d8aba28c4cf8b","abstract_canon_sha256":"040081c2626608e748109eabbf0e1a2ff3bdcd0b6174f33a2f8b4baa8ce8e92c"},"schema_version":"1.0"},"canonical_sha256":"56b745a3bc85bd82050b432c8a7fbaf7bf8a6d8abfb00b9de33962f468d6710c","source":{"kind":"arxiv","id":"2412.13670","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.13670","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"arxiv_version","alias_value":"2412.13670v2","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.13670","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_12","alias_value":"K23ULI54QW6Y","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_16","alias_value":"K23ULI54QW6YEBIL","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_8","alias_value":"K23ULI54","created_at":"2026-07-05T11:11:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:K23ULI54QW6YEBILIMWIU75266","target":"record","payload":{"canonical_record":{"source":{"id":"2412.13670","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-18T09:53:12Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4ebf8ff6c99a7cc4f356bc26440de1bff54163aa8c2b9622272d8aba28c4cf8b","abstract_canon_sha256":"040081c2626608e748109eabbf0e1a2ff3bdcd0b6174f33a2f8b4baa8ce8e92c"},"schema_version":"1.0"},"canonical_sha256":"56b745a3bc85bd82050b432c8a7fbaf7bf8a6d8abfb00b9de33962f468d6710c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:38.629933Z","signature_b64":"fsHS2oY+SyqCwbLQwKqwaaKnb/dO6UL/LfDuUZ5pwAZ7ldeTLqpE54ifP54DZFmTNskTCPKzkYQtLHC/eEDPCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"56b745a3bc85bd82050b432c8a7fbaf7bf8a6d8abfb00b9de33962f468d6710c","last_reissued_at":"2026-07-05T11:11:38.629405Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:38.629405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.13670","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":"2026-07-05T11:11:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"q7Fj+lK6J3SNIvqPB1voQaAM548yt/M53n6z2Q6J4W4c7HJ5dFZnDK8TXJGPmk17dEqkpq8+NipmOsEdwe0RDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T08:05:17.654372Z"},"content_sha256":"4a58ddeacaece47fe3c959af71696b0eef55bb322670eb92302279ce2db48167","schema_version":"1.0","event_id":"sha256:4a58ddeacaece47fe3c959af71696b0eef55bb322670eb92302279ce2db48167"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:K23ULI54QW6YEBILIMWIU75266","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AntiLeakBench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Anh Tuan Luu, Liangming Pan, Mingzhe Du, Rui Mao, Ruiwen Zhou, Shuai Zhao, William Yang Wang, Xiaobao Wu, Yubo Ma, Yuxi Xie","submitted_at":"2024-12-18T09:53:12Z","abstract_excerpt":"Data contamination hinders fair LLM evaluation by introducing test data into newer models' training sets. Existing studies solve this challenge by updating benchmarks with newly collected data. However, they fail to guarantee contamination-free evaluation as the newly collected data may contain pre-existing knowledge, and their benchmark updates rely on intensive human labor. To address these issues, we in this paper propose AntiLeak-Bench, an automated anti-leakage benchmarking framework. Instead of simply using newly collected data, we construct samples with explicitly new knowledge absent f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.13670","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/2412.13670/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":"2026-07-05T11:11:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oAqgKjG4NilD+zKLEOHwTixz+aS52PHlf2HGoQ3X13ZhRkRjDyYmiOWueSIsq5N7pfvnoDeCKzLtfXc1Lv5mAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T08:05:17.654676Z"},"content_sha256":"e73fced54adda12de0d4bc86fbfba67b1593ed1b1e277c74cbcccb898a4eb969","schema_version":"1.0","event_id":"sha256:e73fced54adda12de0d4bc86fbfba67b1593ed1b1e277c74cbcccb898a4eb969"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/K23ULI54QW6YEBILIMWIU75266/bundle.json","state_url":"https://pith.science/pith/K23ULI54QW6YEBILIMWIU75266/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/K23ULI54QW6YEBILIMWIU75266/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":"2026-08-01T08:05:17Z","links":{"resolver":"https://pith.science/pith/K23ULI54QW6YEBILIMWIU75266","bundle":"https://pith.science/pith/K23ULI54QW6YEBILIMWIU75266/bundle.json","state":"https://pith.science/pith/K23ULI54QW6YEBILIMWIU75266/state.json","well_known_bundle":"https://pith.science/.well-known/pith/K23ULI54QW6YEBILIMWIU75266/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:K23ULI54QW6YEBILIMWIU75266","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":"040081c2626608e748109eabbf0e1a2ff3bdcd0b6174f33a2f8b4baa8ce8e92c","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-18T09:53:12Z","title_canon_sha256":"4ebf8ff6c99a7cc4f356bc26440de1bff54163aa8c2b9622272d8aba28c4cf8b"},"schema_version":"1.0","source":{"id":"2412.13670","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.13670","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"arxiv_version","alias_value":"2412.13670v2","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.13670","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_12","alias_value":"K23ULI54QW6Y","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_16","alias_value":"K23ULI54QW6YEBIL","created_at":"2026-07-05T11:11:38Z"},{"alias_kind":"pith_short_8","alias_value":"K23ULI54","created_at":"2026-07-05T11:11:38Z"}],"graph_snapshots":[{"event_id":"sha256:e73fced54adda12de0d4bc86fbfba67b1593ed1b1e277c74cbcccb898a4eb969","target":"graph","created_at":"2026-07-05T11:11:38Z","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/2412.13670/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Data contamination hinders fair LLM evaluation by introducing test data into newer models' training sets. Existing studies solve this challenge by updating benchmarks with newly collected data. However, they fail to guarantee contamination-free evaluation as the newly collected data may contain pre-existing knowledge, and their benchmark updates rely on intensive human labor. To address these issues, we in this paper propose AntiLeak-Bench, an automated anti-leakage benchmarking framework. Instead of simply using newly collected data, we construct samples with explicitly new knowledge absent f","authors_text":"Anh Tuan Luu, Liangming Pan, Mingzhe Du, Rui Mao, Ruiwen Zhou, Shuai Zhao, William Yang Wang, Xiaobao Wu, Yubo Ma, Yuxi Xie","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-18T09:53:12Z","title":"AntiLeakBench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.13670","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":"sha256:4a58ddeacaece47fe3c959af71696b0eef55bb322670eb92302279ce2db48167","target":"record","created_at":"2026-07-05T11:11:38Z","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":"040081c2626608e748109eabbf0e1a2ff3bdcd0b6174f33a2f8b4baa8ce8e92c","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-18T09:53:12Z","title_canon_sha256":"4ebf8ff6c99a7cc4f356bc26440de1bff54163aa8c2b9622272d8aba28c4cf8b"},"schema_version":"1.0","source":{"id":"2412.13670","kind":"arxiv","version":2}},"canonical_sha256":"56b745a3bc85bd82050b432c8a7fbaf7bf8a6d8abfb00b9de33962f468d6710c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"56b745a3bc85bd82050b432c8a7fbaf7bf8a6d8abfb00b9de33962f468d6710c","first_computed_at":"2026-07-05T11:11:38.629405Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:11:38.629405Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fsHS2oY+SyqCwbLQwKqwaaKnb/dO6UL/LfDuUZ5pwAZ7ldeTLqpE54ifP54DZFmTNskTCPKzkYQtLHC/eEDPCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:11:38.629933Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.13670","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4a58ddeacaece47fe3c959af71696b0eef55bb322670eb92302279ce2db48167","sha256:e73fced54adda12de0d4bc86fbfba67b1593ed1b1e277c74cbcccb898a4eb969"],"state_sha256":"bf9d912fd46b43416443da5b118b0ddf5d6fe82cc7fc3603393f6e9f92f87b3a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SfewSqZtrG78vVQTHKU4SkPFWqyqoyqBKWT5sa8W9hYMkf9rLvzx/1CrtnonWL519L4QuGwUh5GwdpjeeNpVCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-01T08:05:17.658942Z","bundle_sha256":"1b96e6478113649e610ac1d45aacbb91dab669e87d5614879ebf5d6c1f7d805f"}}