{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HSXMZTVIETHPLUPTHWF2BDQGDF","short_pith_number":"pith:HSXMZTVI","schema_version":"1.0","canonical_sha256":"3caecccea824cef5d1f33d8ba08e061971d77d68a74509c874c5ab328380c60f","source":{"kind":"arxiv","id":"2412.13647","version":2},"attestation_state":"computed","paper":{"title":"G-VEval: A Versatile Metric for Evaluating Image and Video Captions Using GPT-4o","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Dit-Yan Yeung, Sirui He, Tony Cheng Tong, Zhiwen Shao","submitted_at":"2024-12-18T09:23:12Z","abstract_excerpt":"Evaluation metric of visual captioning is important yet not thoroughly explored. Traditional metrics like BLEU, METEOR, CIDEr, and ROUGE often miss semantic depth, while trained metrics such as CLIP-Score, PAC-S, and Polos are limited in zero-shot scenarios. Advanced Language Model-based metrics also struggle with aligning to nuanced human preferences. To address these issues, we introduce G-VEval, a novel metric inspired by G-Eval and powered by the new GPT-4o. G-VEval uses chain-of-thought reasoning in large multimodal models and supports three modes: reference-free, reference-only, and comb"},"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":"2412.13647","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-18T09:23:12Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"93792171128400e757770d6e5d94be321ba8cfbabc0f10db23645efde285e473","abstract_canon_sha256":"ff359d8994e6c3610194ff94b8ecde08f3339bf33e0873e3c704f160647f76fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:51:40.791171Z","signature_b64":"wagIwsxVtHFL21YElUiu0SqnnKG5I+xuk4lGygS6WEn5B69zOsQF7a1VFRIIf/kUWHd7V2VA7aBjBads1q5HBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3caecccea824cef5d1f33d8ba08e061971d77d68a74509c874c5ab328380c60f","last_reissued_at":"2026-07-05T09:51:40.790680Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:51:40.790680Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"G-VEval: A Versatile Metric for Evaluating Image and Video Captions Using GPT-4o","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Dit-Yan Yeung, Sirui He, Tony Cheng Tong, Zhiwen Shao","submitted_at":"2024-12-18T09:23:12Z","abstract_excerpt":"Evaluation metric of visual captioning is important yet not thoroughly explored. Traditional metrics like BLEU, METEOR, CIDEr, and ROUGE often miss semantic depth, while trained metrics such as CLIP-Score, PAC-S, and Polos are limited in zero-shot scenarios. Advanced Language Model-based metrics also struggle with aligning to nuanced human preferences. To address these issues, we introduce G-VEval, a novel metric inspired by G-Eval and powered by the new GPT-4o. G-VEval uses chain-of-thought reasoning in large multimodal models and supports three modes: reference-free, reference-only, and comb"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.13647","kind":"arxiv","version":2},"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/2412.13647/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":"2412.13647","created_at":"2026-07-05T09:51:40.790742+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.13647v2","created_at":"2026-07-05T09:51:40.790742+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.13647","created_at":"2026-07-05T09:51:40.790742+00:00"},{"alias_kind":"pith_short_12","alias_value":"HSXMZTVIETHP","created_at":"2026-07-05T09:51:40.790742+00:00"},{"alias_kind":"pith_short_16","alias_value":"HSXMZTVIETHPLUPT","created_at":"2026-07-05T09:51:40.790742+00:00"},{"alias_kind":"pith_short_8","alias_value":"HSXMZTVI","created_at":"2026-07-05T09:51:40.790742+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.06600","citing_title":"Evaluation of Multilingual Image Captioning: How far can we get with CLIP models?","ref_index":35,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF","json":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF.json","graph_json":"https://pith.science/api/pith-number/HSXMZTVIETHPLUPTHWF2BDQGDF/graph.json","events_json":"https://pith.science/api/pith-number/HSXMZTVIETHPLUPTHWF2BDQGDF/events.json","paper":"https://pith.science/paper/HSXMZTVI"},"agent_actions":{"view_html":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF","download_json":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF.json","view_paper":"https://pith.science/paper/HSXMZTVI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.13647&json=true","fetch_graph":"https://pith.science/api/pith-number/HSXMZTVIETHPLUPTHWF2BDQGDF/graph.json","fetch_events":"https://pith.science/api/pith-number/HSXMZTVIETHPLUPTHWF2BDQGDF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF/action/storage_attestation","attest_author":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF/action/author_attestation","sign_citation":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF/action/citation_signature","submit_replication":"https://pith.science/pith/HSXMZTVIETHPLUPTHWF2BDQGDF/action/replication_record"}},"created_at":"2026-07-05T09:51:40.790742+00:00","updated_at":"2026-07-05T09:51:40.790742+00:00"}