{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:YR7SR3VO5XL45PGGRWY6QGPB2Z","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":"80f0942eaf7d7ab8b166e5bfbd97250adea70c0b992553cccb09044d68c9bda0","cross_cats_sorted":["cs.CL","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-04-27T16:38:17Z","title_canon_sha256":"083b063049f375cdc06c15d72f47fd3973f592b06bc96d0cbb879a0ebeff5972"},"schema_version":"1.0","source":{"id":"2304.14317","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2304.14317","created_at":"2026-07-05T07:36:00Z"},{"alias_kind":"arxiv_version","alias_value":"2304.14317v2","created_at":"2026-07-05T07:36:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.14317","created_at":"2026-07-05T07:36:00Z"},{"alias_kind":"pith_short_12","alias_value":"YR7SR3VO5XL4","created_at":"2026-07-05T07:36:00Z"},{"alias_kind":"pith_short_16","alias_value":"YR7SR3VO5XL45PGG","created_at":"2026-07-05T07:36:00Z"},{"alias_kind":"pith_short_8","alias_value":"YR7SR3VO","created_at":"2026-07-05T07:36:00Z"}],"graph_snapshots":[{"event_id":"sha256:2486b2f6926dd75eb8928fad68ea2a1cfab2ecec422f20732d36a7ee04e1b702","target":"graph","created_at":"2026-07-05T07:36:00Z","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/2304.14317/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advancements in the field of natural language generation have facilitated the use of large language models to assess the quality of generated text. Although these models have shown promising results in tasks such as machine translation and summarization, their applicability in code intelligence tasks remains limited without human involvement. The complexity of programming concepts required for such tasks makes it difficult to develop evaluation metrics that align with human judgment. Token-matching-based metrics, such as BLEU, have demonstrated weak correlations with human practitioners","authors_text":"Terry Yue Zhuo","cross_cats":["cs.CL","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-04-27T16:38:17Z","title":"ICE-Score: Instructing Large Language Models to Evaluate Code"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.14317","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:c95f3da2857389a4e2d3434c97a16a0442266a33fc2c935e980d92071940ee17","target":"record","created_at":"2026-07-05T07:36:00Z","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":"80f0942eaf7d7ab8b166e5bfbd97250adea70c0b992553cccb09044d68c9bda0","cross_cats_sorted":["cs.CL","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-04-27T16:38:17Z","title_canon_sha256":"083b063049f375cdc06c15d72f47fd3973f592b06bc96d0cbb879a0ebeff5972"},"schema_version":"1.0","source":{"id":"2304.14317","kind":"arxiv","version":2}},"canonical_sha256":"c47f28eeaeedd7cebcc68db1e819e1d667d92269618193534c69773d629525b4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c47f28eeaeedd7cebcc68db1e819e1d667d92269618193534c69773d629525b4","first_computed_at":"2026-07-05T07:36:00.319412Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:36:00.319412Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SgfRyGrNTA0+2YBcd/dDfszZHBTYr25exyBNLRIjGaIRTTPWcHueWTGFbiqsUb+ZRu1t1VUJq7tayFuJsXvzBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:36:00.319869Z","signed_message":"canonical_sha256_bytes"},"source_id":"2304.14317","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c95f3da2857389a4e2d3434c97a16a0442266a33fc2c935e980d92071940ee17","sha256:2486b2f6926dd75eb8928fad68ea2a1cfab2ecec422f20732d36a7ee04e1b702"],"state_sha256":"17629eb454f46f6063c7ed8bf2297e5ce19540f784f6346850f9378861f28e35"}