{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MULX42CSHAP2VXPBWXGHXBJPYX","short_pith_number":"pith:MULX42CS","schema_version":"1.0","canonical_sha256":"65177e6852381faadde1b5cc7b852fc5f0acbb80bf8fb54114b67bc22db4d3fc","source":{"kind":"arxiv","id":"2607.28587","version":1},"attestation_state":"computed","paper":{"title":"PAIChecker: Uncovering and Checking PR-Issue Misalignment in SWE-Bench-Like Benchmarks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Junjielong Xu, Manyi Wang, Pinjia He","submitted_at":"2026-07-30T17:42:44Z","abstract_excerpt":"SWE-bench-like benchmarks are widely used for evaluating LLM's issue resolution capability. They typically follow a common construction pipeline: each PR (Pull Request) is paired with its linked issue by extracting issue references from the PR description; the issue description is used as the problem statement, and the PR patch serves as the test oracle. However, due to the inherent complexity of developing and maintaining large repositories, such PR-Issue pairings are often misaligned in practice. In this work, we systematically study SWE-bench Verified instances, finding that 13.6% exhibit m"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.28587","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-07-30T17:42:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e14ffd0a332635d57fa1b6b4ffcbd5033b19defe15e9fd70b1b28e3016997888","abstract_canon_sha256":"fe28062cb58a0deca7300f9f54bc2193993ab302ebfbd2e054dc26eba0643a93"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"65177e6852381faadde1b5cc7b852fc5f0acbb80bf8fb54114b67bc22db4d3fc","last_reissued_at":"2026-07-31T01:38:47.351479Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:38:47.351479Z"},"graph_snapshot":{"paper":{"title":"PAIChecker: Uncovering and Checking PR-Issue Misalignment in SWE-Bench-Like Benchmarks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Junjielong Xu, Manyi Wang, Pinjia He","submitted_at":"2026-07-30T17:42:44Z","abstract_excerpt":"SWE-bench-like benchmarks are widely used for evaluating LLM's issue resolution capability. They typically follow a common construction pipeline: each PR (Pull Request) is paired with its linked issue by extracting issue references from the PR description; the issue description is used as the problem statement, and the PR patch serves as the test oracle. However, due to the inherent complexity of developing and maintaining large repositories, such PR-Issue pairings are often misaligned in practice. In this work, we systematically study SWE-bench Verified instances, finding that 13.6% exhibit m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.28587","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/2607.28587/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.28587","created_at":"2026-07-31T01:38:47.354727+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.28587v1","created_at":"2026-07-31T01:38:47.354727+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.28587","created_at":"2026-07-31T01:38:47.354727+00:00"},{"alias_kind":"pith_short_12","alias_value":"MULX42CSHAP2","created_at":"2026-07-31T01:38:47.354727+00:00"},{"alias_kind":"pith_short_16","alias_value":"MULX42CSHAP2VXPB","created_at":"2026-07-31T01:38:47.354727+00:00"},{"alias_kind":"pith_short_8","alias_value":"MULX42CS","created_at":"2026-07-31T01:38:47.354727+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX","json":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX.json","graph_json":"https://pith.science/api/pith-number/MULX42CSHAP2VXPBWXGHXBJPYX/graph.json","events_json":"https://pith.science/api/pith-number/MULX42CSHAP2VXPBWXGHXBJPYX/events.json","paper":"https://pith.science/paper/MULX42CS"},"agent_actions":{"view_html":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX","download_json":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX.json","view_paper":"https://pith.science/paper/MULX42CS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.28587&json=true","fetch_graph":"https://pith.science/api/pith-number/MULX42CSHAP2VXPBWXGHXBJPYX/graph.json","fetch_events":"https://pith.science/api/pith-number/MULX42CSHAP2VXPBWXGHXBJPYX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX/action/storage_attestation","attest_author":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX/action/author_attestation","sign_citation":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX/action/citation_signature","submit_replication":"https://pith.science/pith/MULX42CSHAP2VXPBWXGHXBJPYX/action/replication_record"}},"created_at":"2026-07-31T01:38:47.354727+00:00","updated_at":"2026-07-31T01:38:47.354727+00:00"}