{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZQOHKRINA3KGM4K4MAAIE2ACMO","short_pith_number":"pith:ZQOHKRIN","schema_version":"1.0","canonical_sha256":"cc1c75450d06d466715c600082680263afe00b5d2f53b5b133b0e9784e0335cd","source":{"kind":"arxiv","id":"2506.08480","version":1},"attestation_state":"computed","paper":{"title":"Re-Thinking the Automatic Evaluation of Image-Text Alignment in Text-to-Image Models","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Huixuan Zhang, Xiaojun Wan","submitted_at":"2025-06-10T06:11:36Z","abstract_excerpt":"Text-to-image models often struggle to generate images that precisely match textual prompts. Prior research has extensively studied the evaluation of image-text alignment in text-to-image generation. However, existing evaluations primarily focus on agreement with human assessments, neglecting other critical properties of a trustworthy evaluation framework. In this work, we first identify two key aspects that a reliable evaluation should address. We then empirically demonstrate that current mainstream evaluation frameworks fail to fully satisfy these properties across a diverse range of metrics"},"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":"2506.08480","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-10T06:11:36Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"851399c8ba136428458c01ae8819dab0e965f4f2e6326fc3e2e34ee5dbbc9818","abstract_canon_sha256":"ba5669557f73e9b4467f52a3e15c3c9d53fb1ddd2359dd01e782283c14b99a77"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:50.374417Z","signature_b64":"fuUYvuq97Px1w5NvwpcMSdfvWJpRt/DethPy25zJ7b8JVEzsMkwbSCtZKqr49mHxdBsirRpk+HLo5rV5hNF+Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cc1c75450d06d466715c600082680263afe00b5d2f53b5b133b0e9784e0335cd","last_reissued_at":"2026-07-05T11:18:50.373941Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:50.373941Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Re-Thinking the Automatic Evaluation of Image-Text Alignment in Text-to-Image Models","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Huixuan Zhang, Xiaojun Wan","submitted_at":"2025-06-10T06:11:36Z","abstract_excerpt":"Text-to-image models often struggle to generate images that precisely match textual prompts. Prior research has extensively studied the evaluation of image-text alignment in text-to-image generation. However, existing evaluations primarily focus on agreement with human assessments, neglecting other critical properties of a trustworthy evaluation framework. In this work, we first identify two key aspects that a reliable evaluation should address. We then empirically demonstrate that current mainstream evaluation frameworks fail to fully satisfy these properties across a diverse range of metrics"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08480","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/2506.08480/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":"2506.08480","created_at":"2026-07-05T11:18:50.374002+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08480v1","created_at":"2026-07-05T11:18:50.374002+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08480","created_at":"2026-07-05T11:18:50.374002+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZQOHKRINA3KG","created_at":"2026-07-05T11:18:50.374002+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZQOHKRINA3KGM4K4","created_at":"2026-07-05T11:18:50.374002+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZQOHKRIN","created_at":"2026-07-05T11:18:50.374002+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14842","citing_title":"Editor's Choice: Evaluating Abstract Intent in Image Editing through Atomic Entity Analysis","ref_index":70,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO","json":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO.json","graph_json":"https://pith.science/api/pith-number/ZQOHKRINA3KGM4K4MAAIE2ACMO/graph.json","events_json":"https://pith.science/api/pith-number/ZQOHKRINA3KGM4K4MAAIE2ACMO/events.json","paper":"https://pith.science/paper/ZQOHKRIN"},"agent_actions":{"view_html":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO","download_json":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO.json","view_paper":"https://pith.science/paper/ZQOHKRIN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08480&json=true","fetch_graph":"https://pith.science/api/pith-number/ZQOHKRINA3KGM4K4MAAIE2ACMO/graph.json","fetch_events":"https://pith.science/api/pith-number/ZQOHKRINA3KGM4K4MAAIE2ACMO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO/action/storage_attestation","attest_author":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO/action/author_attestation","sign_citation":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO/action/citation_signature","submit_replication":"https://pith.science/pith/ZQOHKRINA3KGM4K4MAAIE2ACMO/action/replication_record"}},"created_at":"2026-07-05T11:18:50.374002+00:00","updated_at":"2026-07-05T11:18:50.374002+00:00"}