{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PY4TZ2EVE5XQ432JROJEHJ46WE","short_pith_number":"pith:PY4TZ2EV","schema_version":"1.0","canonical_sha256":"7e393ce895276f0e6f498b9243a79eb10bfc893d245e9ba3bd095cf0e16920b3","source":{"kind":"arxiv","id":"2406.16851","version":3},"attestation_state":"computed","paper":{"title":"Losing Visual Needles in Image Haystacks: Vision Language Models are Easily Distracted in Short and Long Contexts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Aditya Sharma, Michael Saxon, William Yang Wang","submitted_at":"2024-06-24T17:58:03Z","abstract_excerpt":"We present LoCoVQA, a dynamic benchmark generator for evaluating long-context extractive reasoning in vision language models (VLMs). LoCoVQA augments test examples for mathematical reasoning, VQA, and character recognition tasks with increasingly long visual contexts composed of both in-distribution and out-of-distribution distractor images.\n  Across these tasks, a diverse set of VLMs rapidly lose performance as the visual context length grows, often exhibiting a striking logarithmic decay trend. This test assesses how well VLMs can ignore irrelevant information when answering queries -- a tas"},"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":"2406.16851","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-24T17:58:03Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"c3a5af1999e5bcfd220f4a0db85943a2dda6dbae0f83408ce055a15952c2cb52","abstract_canon_sha256":"686e96ddc2d938814c7113d47ae14af53a1b63b65722c8542fea77a722bb7f03"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:15:34.347766Z","signature_b64":"ZFL0VRTUA8KnMcG1lmOqjLv4hv8lSHVqoAyejgJ0nIcPDZuRDC5PaH1423oW4N0ltM3CPWtrwMYyq8SKJS1wAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e393ce895276f0e6f498b9243a79eb10bfc893d245e9ba3bd095cf0e16920b3","last_reissued_at":"2026-07-05T09:15:34.347214Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:15:34.347214Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Losing Visual Needles in Image Haystacks: Vision Language Models are Easily Distracted in Short and Long Contexts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.CL","authors_text":"Aditya Sharma, Michael Saxon, William Yang Wang","submitted_at":"2024-06-24T17:58:03Z","abstract_excerpt":"We present LoCoVQA, a dynamic benchmark generator for evaluating long-context extractive reasoning in vision language models (VLMs). LoCoVQA augments test examples for mathematical reasoning, VQA, and character recognition tasks with increasingly long visual contexts composed of both in-distribution and out-of-distribution distractor images.\n  Across these tasks, a diverse set of VLMs rapidly lose performance as the visual context length grows, often exhibiting a striking logarithmic decay trend. This test assesses how well VLMs can ignore irrelevant information when answering queries -- a tas"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.16851","kind":"arxiv","version":3},"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/2406.16851/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":"2406.16851","created_at":"2026-07-05T09:15:34.347275+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.16851v3","created_at":"2026-07-05T09:15:34.347275+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.16851","created_at":"2026-07-05T09:15:34.347275+00:00"},{"alias_kind":"pith_short_12","alias_value":"PY4TZ2EVE5XQ","created_at":"2026-07-05T09:15:34.347275+00:00"},{"alias_kind":"pith_short_16","alias_value":"PY4TZ2EVE5XQ432J","created_at":"2026-07-05T09:15:34.347275+00:00"},{"alias_kind":"pith_short_8","alias_value":"PY4TZ2EV","created_at":"2026-07-05T09:15:34.347275+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2504.06925","citing_title":"Are Vision-Language Models Ready for Dietary Assessment? Exploring the Next Frontier in AI-Powered Food Image Recognition","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE","json":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE.json","graph_json":"https://pith.science/api/pith-number/PY4TZ2EVE5XQ432JROJEHJ46WE/graph.json","events_json":"https://pith.science/api/pith-number/PY4TZ2EVE5XQ432JROJEHJ46WE/events.json","paper":"https://pith.science/paper/PY4TZ2EV"},"agent_actions":{"view_html":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE","download_json":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE.json","view_paper":"https://pith.science/paper/PY4TZ2EV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.16851&json=true","fetch_graph":"https://pith.science/api/pith-number/PY4TZ2EVE5XQ432JROJEHJ46WE/graph.json","fetch_events":"https://pith.science/api/pith-number/PY4TZ2EVE5XQ432JROJEHJ46WE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE/action/storage_attestation","attest_author":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE/action/author_attestation","sign_citation":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE/action/citation_signature","submit_replication":"https://pith.science/pith/PY4TZ2EVE5XQ432JROJEHJ46WE/action/replication_record"}},"created_at":"2026-07-05T09:15:34.347275+00:00","updated_at":"2026-07-05T09:15:34.347275+00:00"}