{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:56TB5QA74AYGCRV2Y2S5QSPAV5","short_pith_number":"pith:56TB5QA7","schema_version":"1.0","canonical_sha256":"efa61ec01fe0306146bac6a5d849e0af6a17ad8bb844601aa92128fe65157c4f","source":{"kind":"arxiv","id":"2607.15205","version":1},"attestation_state":"computed","paper":{"title":"MM-IssueLoc: A Controlled Benchmark for Evaluating Visual Evidence in Multimodal Repository-Level Issue Localization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Andrew Gong, Boyu Feng, Hai Lin, Hai-Tao Zheng, Hao Su, Jiaying Zhou, Shaoxiong Zhan, Shi Hu, Yunyun Hou, Zhengda Zhou","submitted_at":"2026-07-16T17:02:25Z","abstract_excerpt":"Real repository issues routinely include visual evidence such as screenshots, error dialogs, rendered UI states, and logs, yet repository-level issue localization is evaluated mostly as a text-only task. Existing multimodal SE benchmarks evaluate end-to-end repair, entangling localization with patch synthesis and obscuring whether visual input helped, hurt, or was ignored. We introduce \\textbf{MM-IssueLoc}, a controlled benchmark and evaluation protocol for repository-level localization with visual evidence. MM-IssueLoc contains 652 issue-PR instances across 23 languages, with annotations for "},"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":"2607.15205","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-07-16T17:02:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"633ac3e3d464088f20343fbb30fbf80f90f4dc809bee68116a16fde73df4e8fd","abstract_canon_sha256":"b2c8dbba9f7f2b417b7328e0f8b2ee61e77902772e2f8d32435aacafa55d50dc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-17T01:22:14.470104Z","signature_b64":"Wy6fqQfo/j6/1+WGzI6F1atV4TVps64Z1IGmBYwYa2C6N9pgNCxvr2QM5u7xh2jhhbPfYEGs6DGrEgqdvBrnCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"efa61ec01fe0306146bac6a5d849e0af6a17ad8bb844601aa92128fe65157c4f","last_reissued_at":"2026-07-17T01:22:14.469219Z","signature_status":"signed_v1","first_computed_at":"2026-07-17T01:22:14.469219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MM-IssueLoc: A Controlled Benchmark for Evaluating Visual Evidence in Multimodal Repository-Level Issue Localization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Andrew Gong, Boyu Feng, Hai Lin, Hai-Tao Zheng, Hao Su, Jiaying Zhou, Shaoxiong Zhan, Shi Hu, Yunyun Hou, Zhengda Zhou","submitted_at":"2026-07-16T17:02:25Z","abstract_excerpt":"Real repository issues routinely include visual evidence such as screenshots, error dialogs, rendered UI states, and logs, yet repository-level issue localization is evaluated mostly as a text-only task. Existing multimodal SE benchmarks evaluate end-to-end repair, entangling localization with patch synthesis and obscuring whether visual input helped, hurt, or was ignored. We introduce \\textbf{MM-IssueLoc}, a controlled benchmark and evaluation protocol for repository-level localization with visual evidence. MM-IssueLoc contains 652 issue-PR instances across 23 languages, with annotations for "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.15205","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/2607.15205/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":"2607.15205","created_at":"2026-07-17T01:22:14.469670+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.15205v1","created_at":"2026-07-17T01:22:14.469670+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.15205","created_at":"2026-07-17T01:22:14.469670+00:00"},{"alias_kind":"pith_short_12","alias_value":"56TB5QA74AYG","created_at":"2026-07-17T01:22:14.469670+00:00"},{"alias_kind":"pith_short_16","alias_value":"56TB5QA74AYGCRV2","created_at":"2026-07-17T01:22:14.469670+00:00"},{"alias_kind":"pith_short_8","alias_value":"56TB5QA7","created_at":"2026-07-17T01:22:14.469670+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/56TB5QA74AYGCRV2Y2S5QSPAV5","json":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5.json","graph_json":"https://pith.science/api/pith-number/56TB5QA74AYGCRV2Y2S5QSPAV5/graph.json","events_json":"https://pith.science/api/pith-number/56TB5QA74AYGCRV2Y2S5QSPAV5/events.json","paper":"https://pith.science/paper/56TB5QA7"},"agent_actions":{"view_html":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5","download_json":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5.json","view_paper":"https://pith.science/paper/56TB5QA7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.15205&json=true","fetch_graph":"https://pith.science/api/pith-number/56TB5QA74AYGCRV2Y2S5QSPAV5/graph.json","fetch_events":"https://pith.science/api/pith-number/56TB5QA74AYGCRV2Y2S5QSPAV5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5/action/storage_attestation","attest_author":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5/action/author_attestation","sign_citation":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5/action/citation_signature","submit_replication":"https://pith.science/pith/56TB5QA74AYGCRV2Y2S5QSPAV5/action/replication_record"}},"created_at":"2026-07-17T01:22:14.469670+00:00","updated_at":"2026-07-17T01:22:14.469670+00:00"}