{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:B7GYC7B6TBKNLGFBP3NZONKJ3J","short_pith_number":"pith:B7GYC7B6","canonical_record":{"source":{"id":"2406.06902","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-11T02:51:17Z","cross_cats_sorted":[],"title_canon_sha256":"f4f97b36e8397bc0040dcef586258fd370c78d6ee3c2e04c34cdd0c37366aeb3","abstract_canon_sha256":"c4a410a7f8a76ace232cdaf2d3c6c5e7d52e7e7573b30504d34f23ae3bd1313a"},"schema_version":"1.0"},"canonical_sha256":"0fcd817c3e9854d598a17edb973549da485247e56891d73027407a73bfb96c91","source":{"kind":"arxiv","id":"2406.06902","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.06902","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"arxiv_version","alias_value":"2406.06902v1","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.06902","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_12","alias_value":"B7GYC7B6TBKN","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_16","alias_value":"B7GYC7B6TBKNLGFB","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_8","alias_value":"B7GYC7B6","created_at":"2026-07-05T08:30:15Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:B7GYC7B6TBKNLGFBP3NZONKJ3J","target":"record","payload":{"canonical_record":{"source":{"id":"2406.06902","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-11T02:51:17Z","cross_cats_sorted":[],"title_canon_sha256":"f4f97b36e8397bc0040dcef586258fd370c78d6ee3c2e04c34cdd0c37366aeb3","abstract_canon_sha256":"c4a410a7f8a76ace232cdaf2d3c6c5e7d52e7e7573b30504d34f23ae3bd1313a"},"schema_version":"1.0"},"canonical_sha256":"0fcd817c3e9854d598a17edb973549da485247e56891d73027407a73bfb96c91","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:30:15.196506Z","signature_b64":"oInKvi2Ca6P7QAtVcS3v4/sfA2Duenqs3EfDE1efkiRAGv2Ab3Ozfe6oPP3L+ID2xTKimudPgJLJI8tVGpdqCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0fcd817c3e9854d598a17edb973549da485247e56891d73027407a73bfb96c91","last_reissued_at":"2026-07-05T08:30:15.195951Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:30:15.195951Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.06902","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":"2026-07-05T08:30:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E2aLsgsemSh7c9R4h9A5cISwIDNDSCjjGNzcYRKsUMC+Rua5dLYan2So4VNmH/vKANhAwk0He5jziBfZh0VBCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T12:00:16.947530Z"},"content_sha256":"bb5a26bb9e36a06b4cb90601f188d304256f9fed44d35c65566f0e09b14e1a26","schema_version":"1.0","event_id":"sha256:bb5a26bb9e36a06b4cb90601f188d304256f9fed44d35c65566f0e09b14e1a26"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:B7GYC7B6TBKNLGFBP3NZONKJ3J","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CodeScore-R: An Automated Robustness Metric for Assessing the FunctionalCorrectness of Code Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Guang Yang, Xiang Chen, Xiangyu Zhang, Yu Zhou","submitted_at":"2024-06-11T02:51:17Z","abstract_excerpt":"Evaluation metrics are crucial in the field of code synthesis. Commonly used code evaluation metrics canbe classified into three types: match-based, semantic-based, and execution-based. Among them, the execution-basedPass@k metric accurately assesses the functionality of predicted code by executing test cases. However, calculatingthis metric requires a significant amount of overhead, necessitating the design of an automated evaluation metric thatcan assess the functionality of predicted code without the need for test cases. Additionally, a good evaluation metricshould be robust, that is the me"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.06902","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/2406.06902/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":"2026-07-05T08:30:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6Wk+ui2vTX+Z5qgzmMr8ynUGAKT2C+Z1wm3OCUEYoKDbFC8PT9zdPU8Ru8m8F5bXM0EvznHnf5Bw4h5YaapnDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T12:00:16.948014Z"},"content_sha256":"ca704397d4ff76b1b2d7c62e392b92896490b5f9efbed655de247737b444c238","schema_version":"1.0","event_id":"sha256:ca704397d4ff76b1b2d7c62e392b92896490b5f9efbed655de247737b444c238"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/bundle.json","state_url":"https://pith.science/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/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":"2026-08-14T12:00:16Z","links":{"resolver":"https://pith.science/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J","bundle":"https://pith.science/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/bundle.json","state":"https://pith.science/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/state.json","well_known_bundle":"https://pith.science/.well-known/pith/B7GYC7B6TBKNLGFBP3NZONKJ3J/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:B7GYC7B6TBKNLGFBP3NZONKJ3J","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":"c4a410a7f8a76ace232cdaf2d3c6c5e7d52e7e7573b30504d34f23ae3bd1313a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-11T02:51:17Z","title_canon_sha256":"f4f97b36e8397bc0040dcef586258fd370c78d6ee3c2e04c34cdd0c37366aeb3"},"schema_version":"1.0","source":{"id":"2406.06902","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.06902","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"arxiv_version","alias_value":"2406.06902v1","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.06902","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_12","alias_value":"B7GYC7B6TBKN","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_16","alias_value":"B7GYC7B6TBKNLGFB","created_at":"2026-07-05T08:30:15Z"},{"alias_kind":"pith_short_8","alias_value":"B7GYC7B6","created_at":"2026-07-05T08:30:15Z"}],"graph_snapshots":[{"event_id":"sha256:ca704397d4ff76b1b2d7c62e392b92896490b5f9efbed655de247737b444c238","target":"graph","created_at":"2026-07-05T08:30:15Z","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/2406.06902/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluation metrics are crucial in the field of code synthesis. Commonly used code evaluation metrics canbe classified into three types: match-based, semantic-based, and execution-based. Among them, the execution-basedPass@k metric accurately assesses the functionality of predicted code by executing test cases. However, calculatingthis metric requires a significant amount of overhead, necessitating the design of an automated evaluation metric thatcan assess the functionality of predicted code without the need for test cases. Additionally, a good evaluation metricshould be robust, that is the me","authors_text":"Guang Yang, Xiang Chen, Xiangyu Zhang, Yu Zhou","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-11T02:51:17Z","title":"CodeScore-R: An Automated Robustness Metric for Assessing the FunctionalCorrectness of Code Synthesis"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.06902","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:bb5a26bb9e36a06b4cb90601f188d304256f9fed44d35c65566f0e09b14e1a26","target":"record","created_at":"2026-07-05T08:30:15Z","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":"c4a410a7f8a76ace232cdaf2d3c6c5e7d52e7e7573b30504d34f23ae3bd1313a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-11T02:51:17Z","title_canon_sha256":"f4f97b36e8397bc0040dcef586258fd370c78d6ee3c2e04c34cdd0c37366aeb3"},"schema_version":"1.0","source":{"id":"2406.06902","kind":"arxiv","version":1}},"canonical_sha256":"0fcd817c3e9854d598a17edb973549da485247e56891d73027407a73bfb96c91","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0fcd817c3e9854d598a17edb973549da485247e56891d73027407a73bfb96c91","first_computed_at":"2026-07-05T08:30:15.195951Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:30:15.195951Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oInKvi2Ca6P7QAtVcS3v4/sfA2Duenqs3EfDE1efkiRAGv2Ab3Ozfe6oPP3L+ID2xTKimudPgJLJI8tVGpdqCg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:30:15.196506Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.06902","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bb5a26bb9e36a06b4cb90601f188d304256f9fed44d35c65566f0e09b14e1a26","sha256:ca704397d4ff76b1b2d7c62e392b92896490b5f9efbed655de247737b444c238"],"state_sha256":"8d94b339d9bcb6b323d5404a3a47c18eb19790f07b57986d13045c895c2abbb3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rsewW5SpLA/ehTbStj7oQ4zRIw2eYLwtEsP/rlh0LYlLagKF/G716nIMuaa2WOk+VKhpr3bKC+bvNJhrJW30Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T12:00:16.952449Z","bundle_sha256":"cbcfc7326dcde5221b4ba353fc4f0c52027cf1038a99612ffd001cc87242509f"}}