{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:M4POQYTBNALD6ZH6UYHFXM3PBM","short_pith_number":"pith:M4POQYTB","canonical_record":{"source":{"id":"2407.14118","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2024-07-19T08:34:30Z","cross_cats_sorted":[],"title_canon_sha256":"14540327d4ccacc50e1db970d233756bb011c41ab4d2bffab8814408c5f1f5f2","abstract_canon_sha256":"065c81e9641dfb5712d9cd19edcfca75228fb5354a5e50a944dfe6fd28c8184a"},"schema_version":"1.0"},"canonical_sha256":"671ee8626168163f64fea60e5bb36f0b133374c147e4c0647876a3725ebe5e1f","source":{"kind":"arxiv","id":"2407.14118","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.14118","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"arxiv_version","alias_value":"2407.14118v1","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.14118","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_12","alias_value":"M4POQYTBNALD","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_16","alias_value":"M4POQYTBNALD6ZH6","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_8","alias_value":"M4POQYTB","created_at":"2026-07-05T08:45:58Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:M4POQYTBNALD6ZH6UYHFXM3PBM","target":"record","payload":{"canonical_record":{"source":{"id":"2407.14118","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2024-07-19T08:34:30Z","cross_cats_sorted":[],"title_canon_sha256":"14540327d4ccacc50e1db970d233756bb011c41ab4d2bffab8814408c5f1f5f2","abstract_canon_sha256":"065c81e9641dfb5712d9cd19edcfca75228fb5354a5e50a944dfe6fd28c8184a"},"schema_version":"1.0"},"canonical_sha256":"671ee8626168163f64fea60e5bb36f0b133374c147e4c0647876a3725ebe5e1f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:58.671705Z","signature_b64":"3lcJlM/IwLx1JDoT59Hui7KdqNO7wcyXUqnqm2RBhuvuGmdlE76XpBHO8ec1sWOPxGprV4JZpJq4XcAm/9a9AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"671ee8626168163f64fea60e5bb36f0b133374c147e4c0647876a3725ebe5e1f","last_reissued_at":"2026-07-05T08:45:58.671063Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:58.671063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.14118","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":"2026-07-05T08:45:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YgnyDD5NWHq9+Gqtm2TK2dDLRrbfWV1RHBAUDeVOCzsrerAYqCjX5g/3zdUjc9mf2xXANdDACEhZl4FYXQzPCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T21:00:26.384363Z"},"content_sha256":"8e731866b1a0e9441da3d1693e9d2d922c1004ecf7989b84d48474f308f8c0b6","schema_version":"1.0","event_id":"sha256:8e731866b1a0e9441da3d1693e9d2d922c1004ecf7989b84d48474f308f8c0b6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:M4POQYTBNALD6ZH6UYHFXM3PBM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Code Generation: Assessing Code LLM Maturity with Postconditions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Fusen He, Juan Zhai, Minxue Pan","submitted_at":"2024-07-19T08:34:30Z","abstract_excerpt":"Most existing code Large Language Model (LLM) benchmarks, e.g., EvalPlus, focus on the code generation tasks. Namely, they contain a natural language description of a problem and ask the LLM to write code to solve the problem. We argue that they do not capture all capabilities needed to assess the quality of a code LLM. In this paper, we propose a code LLM maturity model, based on the postcondition generation problem, to access a more complete set of code LLM capabilities. We choose the postcondition generation problem as it requires the code LLM to understand the code including semantics, nat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.14118","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/2407.14118/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":"2026-07-05T08:45:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fdOfBjtKr3m4WS9UgR5SN57/7sVWGt0HT+5GXqKRa/Z6R1IWfNdNHk/WMK/3RyWPw47jdq6nyHn3ds+Jy+7WBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T21:00:26.384911Z"},"content_sha256":"6bd10a628847b2aa352f54fb4d9e82708abcfc608eb45373c7718180fb1d31bf","schema_version":"1.0","event_id":"sha256:6bd10a628847b2aa352f54fb4d9e82708abcfc608eb45373c7718180fb1d31bf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/bundle.json","state_url":"https://pith.science/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/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":"2026-08-14T21:00:26Z","links":{"resolver":"https://pith.science/pith/M4POQYTBNALD6ZH6UYHFXM3PBM","bundle":"https://pith.science/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/bundle.json","state":"https://pith.science/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/M4POQYTBNALD6ZH6UYHFXM3PBM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:M4POQYTBNALD6ZH6UYHFXM3PBM","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":"065c81e9641dfb5712d9cd19edcfca75228fb5354a5e50a944dfe6fd28c8184a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2024-07-19T08:34:30Z","title_canon_sha256":"14540327d4ccacc50e1db970d233756bb011c41ab4d2bffab8814408c5f1f5f2"},"schema_version":"1.0","source":{"id":"2407.14118","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.14118","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"arxiv_version","alias_value":"2407.14118v1","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.14118","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_12","alias_value":"M4POQYTBNALD","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_16","alias_value":"M4POQYTBNALD6ZH6","created_at":"2026-07-05T08:45:58Z"},{"alias_kind":"pith_short_8","alias_value":"M4POQYTB","created_at":"2026-07-05T08:45:58Z"}],"graph_snapshots":[{"event_id":"sha256:6bd10a628847b2aa352f54fb4d9e82708abcfc608eb45373c7718180fb1d31bf","target":"graph","created_at":"2026-07-05T08:45:58Z","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/2407.14118/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Most existing code Large Language Model (LLM) benchmarks, e.g., EvalPlus, focus on the code generation tasks. Namely, they contain a natural language description of a problem and ask the LLM to write code to solve the problem. We argue that they do not capture all capabilities needed to assess the quality of a code LLM. In this paper, we propose a code LLM maturity model, based on the postcondition generation problem, to access a more complete set of code LLM capabilities. We choose the postcondition generation problem as it requires the code LLM to understand the code including semantics, nat","authors_text":"Fusen He, Juan Zhai, Minxue Pan","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2024-07-19T08:34:30Z","title":"Beyond Code Generation: Assessing Code LLM Maturity with Postconditions"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.14118","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":"sha256:8e731866b1a0e9441da3d1693e9d2d922c1004ecf7989b84d48474f308f8c0b6","target":"record","created_at":"2026-07-05T08:45:58Z","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":"065c81e9641dfb5712d9cd19edcfca75228fb5354a5e50a944dfe6fd28c8184a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2024-07-19T08:34:30Z","title_canon_sha256":"14540327d4ccacc50e1db970d233756bb011c41ab4d2bffab8814408c5f1f5f2"},"schema_version":"1.0","source":{"id":"2407.14118","kind":"arxiv","version":1}},"canonical_sha256":"671ee8626168163f64fea60e5bb36f0b133374c147e4c0647876a3725ebe5e1f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"671ee8626168163f64fea60e5bb36f0b133374c147e4c0647876a3725ebe5e1f","first_computed_at":"2026-07-05T08:45:58.671063Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:45:58.671063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"3lcJlM/IwLx1JDoT59Hui7KdqNO7wcyXUqnqm2RBhuvuGmdlE76XpBHO8ec1sWOPxGprV4JZpJq4XcAm/9a9AA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:45:58.671705Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.14118","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8e731866b1a0e9441da3d1693e9d2d922c1004ecf7989b84d48474f308f8c0b6","sha256:6bd10a628847b2aa352f54fb4d9e82708abcfc608eb45373c7718180fb1d31bf"],"state_sha256":"3d797b44d897953c890415aa30cea9547f16c8daa96d366c26234b1bd8aae068"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"baaOfvseb4dNdr4wGdESvmLbduwO55tW+HH81rcNoMIshD6xIdBs3Qggj64J8cJSsse5flyHvSGfipSk2jRvCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T21:00:26.390455Z","bundle_sha256":"6875c97b7e04f98ca8873295e2533b0c1b69b7c05a0156dc5ad3c4f0998c2e26"}}