{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CENUQAEAZX5H4AIUS35NYHHSTP","short_pith_number":"pith:CENUQAEA","canonical_record":{"source":{"id":"2505.06880","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-11T07:14:30Z","cross_cats_sorted":[],"title_canon_sha256":"b6cf46e154daadaf496f99e18cc9e65277adb54df123174589e1ef7211999b30","abstract_canon_sha256":"f35de95dadb5729f7ecb1c5add092308f946ac9573a7d7463a77392883dc3481"},"schema_version":"1.0"},"canonical_sha256":"111b480080cdfa7e011496fadc1cf29bd521a8f8764e8287505319a02dc29734","source":{"kind":"arxiv","id":"2505.06880","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.06880","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"arxiv_version","alias_value":"2505.06880v1","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06880","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_12","alias_value":"CENUQAEAZX5H","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_16","alias_value":"CENUQAEAZX5H4AIU","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_8","alias_value":"CENUQAEA","created_at":"2026-07-05T11:01:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CENUQAEAZX5H4AIUS35NYHHSTP","target":"record","payload":{"canonical_record":{"source":{"id":"2505.06880","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-11T07:14:30Z","cross_cats_sorted":[],"title_canon_sha256":"b6cf46e154daadaf496f99e18cc9e65277adb54df123174589e1ef7211999b30","abstract_canon_sha256":"f35de95dadb5729f7ecb1c5add092308f946ac9573a7d7463a77392883dc3481"},"schema_version":"1.0"},"canonical_sha256":"111b480080cdfa7e011496fadc1cf29bd521a8f8764e8287505319a02dc29734","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:01:38.337902Z","signature_b64":"ffQs5gyYSwvdatH+d8w5m1tPhth0QucxopJW//5q1mt1VZeYhFN4nku/2b3yZ8YGIFYodMasFOdyrIg5jdM6AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"111b480080cdfa7e011496fadc1cf29bd521a8f8764e8287505319a02dc29734","last_reissued_at":"2026-07-05T11:01:38.337406Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:01:38.337406Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.06880","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":"2026-07-05T11:01:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XHxFHpOlzrEu/pB3x9QDc2UbOFIFV7QplR7/r+klxEwgkY7oJsOjkvw1S+khWp4REy+y3v1sxbuGMsur8RbeDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:20:27.695953Z"},"content_sha256":"86de2fc5b5896e73c8cfceb7ca92b23cd559b0cc830864a95ea235d9ceec8fe9","schema_version":"1.0","event_id":"sha256:86de2fc5b5896e73c8cfceb7ca92b23cd559b0cc830864a95ea235d9ceec8fe9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CENUQAEAZX5H4AIUS35NYHHSTP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Benchmarking and Revisiting Code Generation Assessment: A Mutation-Based Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Chao Shen, Jian Wang, Longtian Wang, Tianlin Li, Xiaofei Xie, Yuhan Zhi","submitted_at":"2025-05-11T07:14:30Z","abstract_excerpt":"Code Large Language Models (CLLMs) have exhibited outstanding performance in program synthesis, attracting the focus of the research community. The evaluation of CLLM's program synthesis capability has generally relied on manually curated benchmarks. However, there is a substantial gap between real-world scenarios and benchmark settings. Existing benchmarks typically provide only a single input prompt for the evaluation of each synthesis problem. However, in practice, a problem can be described in various ways, including with typos, where developers may struggle to understand certain descripti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06880","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/2505.06880/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":"2026-07-05T11:01:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tB7QaxtRpmjmMKl0gaZvOY7Jl1bWt1j5Lgc7tePgZEk1nN9bCjhCCMfgM9hmlZ35s7YUeZouqjTLCCrNPWmbCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:20:27.700168Z"},"content_sha256":"12300409ec5dc5935335538f901d2a09f48db1df1e25a7a1eaf4a2122d5df794","schema_version":"1.0","event_id":"sha256:12300409ec5dc5935335538f901d2a09f48db1df1e25a7a1eaf4a2122d5df794"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CENUQAEAZX5H4AIUS35NYHHSTP/bundle.json","state_url":"https://pith.science/pith/CENUQAEAZX5H4AIUS35NYHHSTP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CENUQAEAZX5H4AIUS35NYHHSTP/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":"2026-08-05T13:20:27Z","links":{"resolver":"https://pith.science/pith/CENUQAEAZX5H4AIUS35NYHHSTP","bundle":"https://pith.science/pith/CENUQAEAZX5H4AIUS35NYHHSTP/bundle.json","state":"https://pith.science/pith/CENUQAEAZX5H4AIUS35NYHHSTP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CENUQAEAZX5H4AIUS35NYHHSTP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CENUQAEAZX5H4AIUS35NYHHSTP","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":"f35de95dadb5729f7ecb1c5add092308f946ac9573a7d7463a77392883dc3481","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-11T07:14:30Z","title_canon_sha256":"b6cf46e154daadaf496f99e18cc9e65277adb54df123174589e1ef7211999b30"},"schema_version":"1.0","source":{"id":"2505.06880","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.06880","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"arxiv_version","alias_value":"2505.06880v1","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06880","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_12","alias_value":"CENUQAEAZX5H","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_16","alias_value":"CENUQAEAZX5H4AIU","created_at":"2026-07-05T11:01:38Z"},{"alias_kind":"pith_short_8","alias_value":"CENUQAEA","created_at":"2026-07-05T11:01:38Z"}],"graph_snapshots":[{"event_id":"sha256:12300409ec5dc5935335538f901d2a09f48db1df1e25a7a1eaf4a2122d5df794","target":"graph","created_at":"2026-07-05T11:01:38Z","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/2505.06880/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Code Large Language Models (CLLMs) have exhibited outstanding performance in program synthesis, attracting the focus of the research community. The evaluation of CLLM's program synthesis capability has generally relied on manually curated benchmarks. However, there is a substantial gap between real-world scenarios and benchmark settings. Existing benchmarks typically provide only a single input prompt for the evaluation of each synthesis problem. However, in practice, a problem can be described in various ways, including with typos, where developers may struggle to understand certain descripti","authors_text":"Chao Shen, Jian Wang, Longtian Wang, Tianlin Li, Xiaofei Xie, Yuhan Zhi","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-11T07:14:30Z","title":"Benchmarking and Revisiting Code Generation Assessment: A Mutation-Based Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06880","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":"sha256:86de2fc5b5896e73c8cfceb7ca92b23cd559b0cc830864a95ea235d9ceec8fe9","target":"record","created_at":"2026-07-05T11:01:38Z","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":"f35de95dadb5729f7ecb1c5add092308f946ac9573a7d7463a77392883dc3481","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-11T07:14:30Z","title_canon_sha256":"b6cf46e154daadaf496f99e18cc9e65277adb54df123174589e1ef7211999b30"},"schema_version":"1.0","source":{"id":"2505.06880","kind":"arxiv","version":1}},"canonical_sha256":"111b480080cdfa7e011496fadc1cf29bd521a8f8764e8287505319a02dc29734","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"111b480080cdfa7e011496fadc1cf29bd521a8f8764e8287505319a02dc29734","first_computed_at":"2026-07-05T11:01:38.337406Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:01:38.337406Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ffQs5gyYSwvdatH+d8w5m1tPhth0QucxopJW//5q1mt1VZeYhFN4nku/2b3yZ8YGIFYodMasFOdyrIg5jdM6AA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:01:38.337902Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.06880","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:86de2fc5b5896e73c8cfceb7ca92b23cd559b0cc830864a95ea235d9ceec8fe9","sha256:12300409ec5dc5935335538f901d2a09f48db1df1e25a7a1eaf4a2122d5df794"],"state_sha256":"be061f3d60a64ca9917c0c78cbc5b2b064404c6af904c6e1656de7836a0e9e7f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TgJpfGjFPhwsxEzJ8BkUTqQrHJkihBnpjPJ96k9wp3qnlUk/pRrGspMHxwK2HY0zsUV2CqhfGuRu/Hd52DUgBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T13:20:27.716163Z","bundle_sha256":"6a72c21bb65bacfd72bc6c7adacfc22b2c0d2b553283c801632df74b46cd6d18"}}