{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QM3YSB6RVXIGN6ACWEIDN66QJA","short_pith_number":"pith:QM3YSB6R","schema_version":"1.0","canonical_sha256":"83378907d1add066f802b11036fbd0481f7473ac77fed2c79a5f60b17c321916","source":{"kind":"arxiv","id":"2310.05861","version":2},"attestation_state":"computed","paper":{"title":"Rephrase, Augment, Reason: Visual Grounding of Questions for Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Archiki Prasad, Elias Stengel-Eskin, Mohit Bansal","submitted_at":"2023-10-09T16:57:57Z","abstract_excerpt":"An increasing number of vision-language tasks can be handled with little to no training, i.e., in a zero and few-shot manner, by marrying large language models (LLMs) to vision encoders, resulting in large vision-language models (LVLMs). While this has huge upsides, such as not requiring training data or custom architectures, how an input is presented to an LVLM can have a major impact on zero-shot model performance. In particular, inputs phrased in an underspecified way can result in incorrect answers due to factors like missing visual information, complex implicit reasoning, or linguistic am"},"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":"2310.05861","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-09T16:57:57Z","cross_cats_sorted":["cs.AI","cs.CV","cs.LG"],"title_canon_sha256":"a2b1f8bc1103df688ec2428a4b06c252d8799fce256eb9050b695be26712ad98","abstract_canon_sha256":"414aea8cd130c28301b895eba5427aeca246b21d4eba3adf4abb6e66fb1b5af3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:03:27.163205Z","signature_b64":"msuvqffiIO8XGrRznG+J2qNLM3LH/yK/DZuUXaZtyHHu1xVgfBdhjLwq+CF0VVN9R1ixW17UEYceHKLIouJ7Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83378907d1add066f802b11036fbd0481f7473ac77fed2c79a5f60b17c321916","last_reissued_at":"2026-07-05T08:03:27.162760Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:03:27.162760Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rephrase, Augment, Reason: Visual Grounding of Questions for Vision-Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Archiki Prasad, Elias Stengel-Eskin, Mohit Bansal","submitted_at":"2023-10-09T16:57:57Z","abstract_excerpt":"An increasing number of vision-language tasks can be handled with little to no training, i.e., in a zero and few-shot manner, by marrying large language models (LLMs) to vision encoders, resulting in large vision-language models (LVLMs). While this has huge upsides, such as not requiring training data or custom architectures, how an input is presented to an LVLM can have a major impact on zero-shot model performance. In particular, inputs phrased in an underspecified way can result in incorrect answers due to factors like missing visual information, complex implicit reasoning, or linguistic am"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.05861","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/2310.05861/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":"2310.05861","created_at":"2026-07-05T08:03:27.162811+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.05861v2","created_at":"2026-07-05T08:03:27.162811+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.05861","created_at":"2026-07-05T08:03:27.162811+00:00"},{"alias_kind":"pith_short_12","alias_value":"QM3YSB6RVXIG","created_at":"2026-07-05T08:03:27.162811+00:00"},{"alias_kind":"pith_short_16","alias_value":"QM3YSB6RVXIGN6AC","created_at":"2026-07-05T08:03:27.162811+00:00"},{"alias_kind":"pith_short_8","alias_value":"QM3YSB6R","created_at":"2026-07-05T08:03:27.162811+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.02201","citing_title":"Acknowledging Focus Ambiguity in Visual Questions","ref_index":40,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA","json":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA.json","graph_json":"https://pith.science/api/pith-number/QM3YSB6RVXIGN6ACWEIDN66QJA/graph.json","events_json":"https://pith.science/api/pith-number/QM3YSB6RVXIGN6ACWEIDN66QJA/events.json","paper":"https://pith.science/paper/QM3YSB6R"},"agent_actions":{"view_html":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA","download_json":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA.json","view_paper":"https://pith.science/paper/QM3YSB6R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.05861&json=true","fetch_graph":"https://pith.science/api/pith-number/QM3YSB6RVXIGN6ACWEIDN66QJA/graph.json","fetch_events":"https://pith.science/api/pith-number/QM3YSB6RVXIGN6ACWEIDN66QJA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA/action/storage_attestation","attest_author":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA/action/author_attestation","sign_citation":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA/action/citation_signature","submit_replication":"https://pith.science/pith/QM3YSB6RVXIGN6ACWEIDN66QJA/action/replication_record"}},"created_at":"2026-07-05T08:03:27.162811+00:00","updated_at":"2026-07-05T08:03:27.162811+00:00"}