{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:WE4OYEKYLCLORRKLUYZHPMMI42","short_pith_number":"pith:WE4OYEKY","canonical_record":{"source":{"id":"2603.04459","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-03-03T09:10:45Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"54c10d94cfb6db6425fef0dd3dd4b2613d353ba3c62e0cd7a120b2f4e03a8327","abstract_canon_sha256":"c2fd940e6200b3d207449f08926ca7b45d15068e38a1661f852d2808c9bae8fb"},"schema_version":"1.0"},"canonical_sha256":"b138ec11585896e8c54ba63277b188e6affeeb688111c1457e06ce27c49df270","source":{"kind":"arxiv","id":"2603.04459","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.04459","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"arxiv_version","alias_value":"2603.04459v3","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.04459","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_12","alias_value":"WE4OYEKYLCLO","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_16","alias_value":"WE4OYEKYLCLORRKL","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_8","alias_value":"WE4OYEKY","created_at":"2026-05-20T00:00:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:WE4OYEKYLCLORRKLUYZHPMMI42","target":"record","payload":{"canonical_record":{"source":{"id":"2603.04459","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-03-03T09:10:45Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"54c10d94cfb6db6425fef0dd3dd4b2613d353ba3c62e0cd7a120b2f4e03a8327","abstract_canon_sha256":"c2fd940e6200b3d207449f08926ca7b45d15068e38a1661f852d2808c9bae8fb"},"schema_version":"1.0"},"canonical_sha256":"b138ec11585896e8c54ba63277b188e6affeeb688111c1457e06ce27c49df270","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:35.700199Z","signature_b64":"zNsoz2SnW37tBF1rSjmzLxVI2UOMU3h8JkETojuNVzOSBV4vdTDlOMRuNg9Nk13qwusK5mzVwRojji94mbj5Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b138ec11585896e8c54ba63277b188e6affeeb688111c1457e06ce27c49df270","last_reissued_at":"2026-05-20T00:00:35.699596Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:35.699596Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2603.04459","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:00:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aTEK8/q33IupeB1U0bR6MVTTrxOgjjiFFP/Q4An6xkYqE3oXdmyBHZ1rbVULX0QvIXYeK5UIDDga0MVXHrNAAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T12:04:43.321908Z"},"content_sha256":"a494a798804da7edc7db8d3e593306eb14def9858dfb19cc91807c2f2a561b5f","schema_version":"1.0","event_id":"sha256:a494a798804da7edc7db8d3e593306eb14def9858dfb19cc91807c2f2a561b5f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:WE4OYEKYLCLORRKLUYZHPMMI42","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.CR","authors_text":"Junjie Chu, Michael Backes, Xinyue Shen, Yang Zhang, Ye Leng, Yun Shen","submitted_at":"2026-03-03T09:10:45Z","abstract_excerpt":"The rapid expansion of research in LLM safety presents challenges in tracking advancements, making benchmarks important evaluation infrastructures for identifying key trends and facilitating systematic comparisons. Yet no systematic assessment exists of their code quality and runnability, nor of what factors are associated with the community's adoption of certain benchmarks over others. To address this gap, we conduct a systematic measurement study of 31 LLM safety benchmarks (covering prompt injection, jailbreak, and hallucination) with 382 non-benchmark papers as a control group, combining a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.04459","kind":"arxiv","version":3},"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/2603.04459/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":"2026-05-20T00:00:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4iDfKyRhUch4gOOUoNFy6GN1lUDXmZoAyaYq70S1LaXxMGaiwN54c9h3chFr5fe+PxGGy0ypyh2Y+FD8FbKjDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T12:04:43.322722Z"},"content_sha256":"ea8a14160234c67811151ea04f3d8568c86984c44da36a66e64c33aae319de9e","schema_version":"1.0","event_id":"sha256:ea8a14160234c67811151ea04f3d8568c86984c44da36a66e64c33aae319de9e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WE4OYEKYLCLORRKLUYZHPMMI42/bundle.json","state_url":"https://pith.science/pith/WE4OYEKYLCLORRKLUYZHPMMI42/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WE4OYEKYLCLORRKLUYZHPMMI42/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":"2026-05-25T12:04:43Z","links":{"resolver":"https://pith.science/pith/WE4OYEKYLCLORRKLUYZHPMMI42","bundle":"https://pith.science/pith/WE4OYEKYLCLORRKLUYZHPMMI42/bundle.json","state":"https://pith.science/pith/WE4OYEKYLCLORRKLUYZHPMMI42/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WE4OYEKYLCLORRKLUYZHPMMI42/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:WE4OYEKYLCLORRKLUYZHPMMI42","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":"c2fd940e6200b3d207449f08926ca7b45d15068e38a1661f852d2808c9bae8fb","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-03-03T09:10:45Z","title_canon_sha256":"54c10d94cfb6db6425fef0dd3dd4b2613d353ba3c62e0cd7a120b2f4e03a8327"},"schema_version":"1.0","source":{"id":"2603.04459","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.04459","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"arxiv_version","alias_value":"2603.04459v3","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.04459","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_12","alias_value":"WE4OYEKYLCLO","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_16","alias_value":"WE4OYEKYLCLORRKL","created_at":"2026-05-20T00:00:35Z"},{"alias_kind":"pith_short_8","alias_value":"WE4OYEKY","created_at":"2026-05-20T00:00:35Z"}],"graph_snapshots":[{"event_id":"sha256:ea8a14160234c67811151ea04f3d8568c86984c44da36a66e64c33aae319de9e","target":"graph","created_at":"2026-05-20T00:00:35Z","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/2603.04459/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The rapid expansion of research in LLM safety presents challenges in tracking advancements, making benchmarks important evaluation infrastructures for identifying key trends and facilitating systematic comparisons. Yet no systematic assessment exists of their code quality and runnability, nor of what factors are associated with the community's adoption of certain benchmarks over others. To address this gap, we conduct a systematic measurement study of 31 LLM safety benchmarks (covering prompt injection, jailbreak, and hallucination) with 382 non-benchmark papers as a control group, combining a","authors_text":"Junjie Chu, Michael Backes, Xinyue Shen, Yang Zhang, Ye Leng, Yun Shen","cross_cats":["cs.AI","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-03-03T09:10:45Z","title":"Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.04459","kind":"arxiv","version":3},"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:a494a798804da7edc7db8d3e593306eb14def9858dfb19cc91807c2f2a561b5f","target":"record","created_at":"2026-05-20T00:00:35Z","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":"c2fd940e6200b3d207449f08926ca7b45d15068e38a1661f852d2808c9bae8fb","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-03-03T09:10:45Z","title_canon_sha256":"54c10d94cfb6db6425fef0dd3dd4b2613d353ba3c62e0cd7a120b2f4e03a8327"},"schema_version":"1.0","source":{"id":"2603.04459","kind":"arxiv","version":3}},"canonical_sha256":"b138ec11585896e8c54ba63277b188e6affeeb688111c1457e06ce27c49df270","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b138ec11585896e8c54ba63277b188e6affeeb688111c1457e06ce27c49df270","first_computed_at":"2026-05-20T00:00:35.699596Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:00:35.699596Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zNsoz2SnW37tBF1rSjmzLxVI2UOMU3h8JkETojuNVzOSBV4vdTDlOMRuNg9Nk13qwusK5mzVwRojji94mbj5Dw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:00:35.700199Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.04459","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a494a798804da7edc7db8d3e593306eb14def9858dfb19cc91807c2f2a561b5f","sha256:ea8a14160234c67811151ea04f3d8568c86984c44da36a66e64c33aae319de9e"],"state_sha256":"d39e4b36e0b432911eaab91c5292d9d84fcd2656fdd1c7b87c11b7ac8636c461"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rK6sPZvAf9WC01BLN6phbD74vOxZAomtv6xta5OGcvh1kBKgJ0EjDcS4IZjqTOahuXpU8CCqyokkn5ReYPQyAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T12:04:43.326689Z","bundle_sha256":"822ade04e379adb442b1ef356919378e3c15a9da25553a216cc0e2e57eee5163"}}