{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:FYCPYWUSTT24IKQKWQ7XGJHKXD","short_pith_number":"pith:FYCPYWUS","canonical_record":{"source":{"id":"2605.02096","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-03T23:31:18Z","cross_cats_sorted":[],"title_canon_sha256":"f3465e4af18f27521b9a50ca1af6dc8d470ca4d2a250056bde88bbdf2b113830","abstract_canon_sha256":"aa9076888c1339e94059a74c0c978245ab9e2e3fbcfa4f3da5eb0a43d826cb00"},"schema_version":"1.0"},"canonical_sha256":"2e04fc5a929cf5c42a0ab43f7324eab8d522c9f5c63ee2b9527b036632b09d4e","source":{"kind":"arxiv","id":"2605.02096","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.02096","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"arxiv_version","alias_value":"2605.02096v2","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.02096","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_12","alias_value":"FYCPYWUSTT24","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_16","alias_value":"FYCPYWUSTT24IKQK","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_8","alias_value":"FYCPYWUS","created_at":"2026-07-03T00:16:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:FYCPYWUSTT24IKQKWQ7XGJHKXD","target":"record","payload":{"canonical_record":{"source":{"id":"2605.02096","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-03T23:31:18Z","cross_cats_sorted":[],"title_canon_sha256":"f3465e4af18f27521b9a50ca1af6dc8d470ca4d2a250056bde88bbdf2b113830","abstract_canon_sha256":"aa9076888c1339e94059a74c0c978245ab9e2e3fbcfa4f3da5eb0a43d826cb00"},"schema_version":"1.0"},"canonical_sha256":"2e04fc5a929cf5c42a0ab43f7324eab8d522c9f5c63ee2b9527b036632b09d4e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-03T00:16:53.951922Z","signature_b64":"Q68WG1JyYTvUob2cPwDGUCs+JmKKFZMm49NEGPDFElabEEEHqwSpK1tUPrmMANk8Km/4e8+klgyroV0Xjn79Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e04fc5a929cf5c42a0ab43f7324eab8d522c9f5c63ee2b9527b036632b09d4e","last_reissued_at":"2026-07-03T00:16:53.951432Z","signature_status":"signed_v1","first_computed_at":"2026-07-03T00:16:53.951432Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.02096","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-03T00:16:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"72RZU8rdpq8Kaurp9rgXVPIVbVEfmqs4jh+x2Gaf4zrToGrFCXU09kX1PRDU0JkWW9joJXQ3+watOA932PjLBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T11:56:56.868990Z"},"content_sha256":"a626f3053422c0090dc7bef22dfc5946c87515155b1fda25bb129a9955942b03","schema_version":"1.0","event_id":"sha256:a626f3053422c0090dc7bef22dfc5946c87515155b1fda25bb129a9955942b03"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:FYCPYWUSTT24IKQKWQ7XGJHKXD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Foundation Models as Oracles for Refactoring Correctness Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests.","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Baldoino Fonseca, Jonhnanthan Oliveira, Marcio Ribeiro, Rian Melo, Rohit Gheyi","submitted_at":"2026-05-03T23:31:18Z","abstract_excerpt":"Refactoring tools in popular Integrated Development Environments (IDEs) can introduce unintended behavioral changes or compilation errors, a persistent challenge that undermines developer trust in automated transformations. Traditional detection approaches rely on handcrafted preconditions, and static and dynamic analyses, yet remain limited in adaptability and can miss subtle correctness issues. This study examines the potential of foundation models to serve as oracles for detecting refactoring bugs in Java programs. We evaluate zero-shot prompting, without task-specific training, across 226 "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our results indicate that foundation models can be effective for this task, although performance varies across models. In the first-run setting, GPT-OSS-20B achieved 80.5% accuracy, while GPT-5.4 reached 93.8%.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the foundation models are reasoning about the semantics of the code changes rather than relying on superficial patterns or training data leakage, as partially addressed by metamorphic testing.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Foundation models serve as effective oracles for detecting refactoring correctness issues in Java programs, achieving up to 93.8% accuracy in zero-shot evaluations on 226 real bugs.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"cd44602f7e0e86d1e36bdfe98089f9f60fefddaede5caa06673cfac0947aeaf3"},"source":{"id":"2605.02096","kind":"arxiv","version":2},"verdict":{"id":"b7feacbc-2592-45e9-867b-388227ab5302","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T19:07:02.912635Z","strongest_claim":"Our results indicate that foundation models can be effective for this task, although performance varies across models. In the first-run setting, GPT-OSS-20B achieved 80.5% accuracy, while GPT-5.4 reached 93.8%.","one_line_summary":"Foundation models serve as effective oracles for detecting refactoring correctness issues in Java programs, achieving up to 93.8% accuracy in zero-shot evaluations on 226 real bugs.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the foundation models are reasoning about the semantics of the code changes rather than relying on superficial patterns or training data leakage, as partially addressed by metamorphic testing.","pith_extraction_headline":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.02096/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T16:37:54.503996Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-20T04:01:22.747741Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T16:40:33.145882Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"d9394d9362acf701735b1fb00423ad36c159965394e5613d9bf38ba5c9620b9a"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"747e6eeee80fc46bd85bd201b5aaff9ff648cf103324f2fd6852e7868585f370"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"b7feacbc-2592-45e9-867b-388227ab5302"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-03T00:16:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ulIpkKV62v5er+rMRfK8PzoD6z1JntiO+rNXp0AEjl04y9wfmDtvkyr+9KQwjIswNtEU5OhTMnQ9bYOGblaXAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T11:56:56.869705Z"},"content_sha256":"9fe64af3686177110cda18d7aa33f116aa5f94aaf017fd82d2cf2d403ee407fe","schema_version":"1.0","event_id":"sha256:9fe64af3686177110cda18d7aa33f116aa5f94aaf017fd82d2cf2d403ee407fe"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/bundle.json","state_url":"https://pith.science/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/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":"2026-08-20T11:56:56Z","links":{"resolver":"https://pith.science/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD","bundle":"https://pith.science/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/bundle.json","state":"https://pith.science/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FYCPYWUSTT24IKQKWQ7XGJHKXD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FYCPYWUSTT24IKQKWQ7XGJHKXD","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":"aa9076888c1339e94059a74c0c978245ab9e2e3fbcfa4f3da5eb0a43d826cb00","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-03T23:31:18Z","title_canon_sha256":"f3465e4af18f27521b9a50ca1af6dc8d470ca4d2a250056bde88bbdf2b113830"},"schema_version":"1.0","source":{"id":"2605.02096","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.02096","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"arxiv_version","alias_value":"2605.02096v2","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.02096","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_12","alias_value":"FYCPYWUSTT24","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_16","alias_value":"FYCPYWUSTT24IKQK","created_at":"2026-07-03T00:16:53Z"},{"alias_kind":"pith_short_8","alias_value":"FYCPYWUS","created_at":"2026-07-03T00:16:53Z"}],"graph_snapshots":[{"event_id":"sha256:9fe64af3686177110cda18d7aa33f116aa5f94aaf017fd82d2cf2d403ee407fe","target":"graph","created_at":"2026-07-03T00:16:53Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Our results indicate that foundation models can be effective for this task, although performance varies across models. In the first-run setting, GPT-OSS-20B achieved 80.5% accuracy, while GPT-5.4 reached 93.8%."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the foundation models are reasoning about the semantics of the code changes rather than relying on superficial patterns or training data leakage, as partially addressed by metamorphic testing."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Foundation models serve as effective oracles for detecting refactoring correctness issues in Java programs, achieving up to 93.8% accuracy in zero-shot evaluations on 226 real bugs."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests."}],"snapshot_sha256":"cd44602f7e0e86d1e36bdfe98089f9f60fefddaede5caa06673cfac0947aeaf3"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"747e6eeee80fc46bd85bd201b5aaff9ff648cf103324f2fd6852e7868585f370"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T16:37:54.503996Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-20T04:01:22.747741Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T16:40:33.145882Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.02096/integrity.json","findings":[],"snapshot_sha256":"d9394d9362acf701735b1fb00423ad36c159965394e5613d9bf38ba5c9620b9a","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Refactoring tools in popular Integrated Development Environments (IDEs) can introduce unintended behavioral changes or compilation errors, a persistent challenge that undermines developer trust in automated transformations. Traditional detection approaches rely on handcrafted preconditions, and static and dynamic analyses, yet remain limited in adaptability and can miss subtle correctness issues. This study examines the potential of foundation models to serve as oracles for detecting refactoring bugs in Java programs. We evaluate zero-shot prompting, without task-specific training, across 226 ","authors_text":"Baldoino Fonseca, Jonhnanthan Oliveira, Marcio Ribeiro, Rian Melo, Rohit Gheyi","cross_cats":[],"headline":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-03T23:31:18Z","title":"Foundation Models as Oracles for Refactoring Correctness Detection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.02096","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-08T19:07:02.912635Z","id":"b7feacbc-2592-45e9-867b-388227ab5302","model_set":{"reader":"grok-4.3"},"one_line_summary":"Foundation models serve as effective oracles for detecting refactoring correctness issues in Java programs, achieving up to 93.8% accuracy in zero-shot evaluations on 226 real bugs.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Foundation models can act as oracles to detect when Java refactoring transformations introduce bugs, reaching over 90 percent accuracy in zero-shot tests.","strongest_claim":"Our results indicate that foundation models can be effective for this task, although performance varies across models. In the first-run setting, GPT-OSS-20B achieved 80.5% accuracy, while GPT-5.4 reached 93.8%.","weakest_assumption":"That the foundation models are reasoning about the semantics of the code changes rather than relying on superficial patterns or training data leakage, as partially addressed by metamorphic testing."}},"verdict_id":"b7feacbc-2592-45e9-867b-388227ab5302"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a626f3053422c0090dc7bef22dfc5946c87515155b1fda25bb129a9955942b03","target":"record","created_at":"2026-07-03T00:16:53Z","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":"aa9076888c1339e94059a74c0c978245ab9e2e3fbcfa4f3da5eb0a43d826cb00","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-03T23:31:18Z","title_canon_sha256":"f3465e4af18f27521b9a50ca1af6dc8d470ca4d2a250056bde88bbdf2b113830"},"schema_version":"1.0","source":{"id":"2605.02096","kind":"arxiv","version":2}},"canonical_sha256":"2e04fc5a929cf5c42a0ab43f7324eab8d522c9f5c63ee2b9527b036632b09d4e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2e04fc5a929cf5c42a0ab43f7324eab8d522c9f5c63ee2b9527b036632b09d4e","first_computed_at":"2026-07-03T00:16:53.951432Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-03T00:16:53.951432Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Q68WG1JyYTvUob2cPwDGUCs+JmKKFZMm49NEGPDFElabEEEHqwSpK1tUPrmMANk8Km/4e8+klgyroV0Xjn79Bw==","signature_status":"signed_v1","signed_at":"2026-07-03T00:16:53.951922Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.02096","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a626f3053422c0090dc7bef22dfc5946c87515155b1fda25bb129a9955942b03","sha256:9fe64af3686177110cda18d7aa33f116aa5f94aaf017fd82d2cf2d403ee407fe"],"state_sha256":"16f233bc954085efc7e3138ffe775d3372bd11fbded447621ccfc9503c207873"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ecyK16TIrIj8BDLU+aOzphVokWNcpHvsgpw4QqIJ6LRuf9josT7OhSMudrmo8xn0SvfTMVYbXmKRq1qJv3nQBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T11:56:56.874962Z","bundle_sha256":"075e6b64d635ff9c54352be782064c6eb1f78b974bbf7178d0b77ac13dcd3cfb"}}