{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HWKGJ7SMZBKI4RSN64HTJWAFE2","short_pith_number":"pith:HWKGJ7SM","schema_version":"1.0","canonical_sha256":"3d9464fe4cc8548e464df70f34d80526a0ae630b90a2a55883131b676d750d00","source":{"kind":"arxiv","id":"2505.08971","version":1},"attestation_state":"computed","paper":{"title":"Prioritizing Image-Related Tokens Enhances Vision-Language Pre-Training","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hao Peng, Heng Ji, Tong Zhang, Yangyi Chen","submitted_at":"2025-05-13T21:27:52Z","abstract_excerpt":"In standard large vision-language models (LVLMs) pre-training, the model typically maximizes the joint probability of the caption conditioned on the image via next-token prediction (NTP); however, since only a small subset of caption tokens directly relates to the visual content, this naive NTP unintentionally fits the model to noise and increases the risk of hallucination. We present PRIOR, a simple vision-language pre-training approach that addresses this issue by prioritizing image-related tokens through differential weighting in the NTP loss, drawing from the importance sampling framework."},"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":"2505.08971","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-05-13T21:27:52Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"b224f882801ce7020111049adadfd80e49372b2c8635a66989ef13a77a27e25f","abstract_canon_sha256":"bf6d20e31cea3518a011a66927b53177e2e4613277cb749dce9bac898259e129"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:02:48.130242Z","signature_b64":"kbnOW6pf2MUW/zNekW6hUFktnm0rdLCaRnijJoG0OQEEG62JuNc5ZlkIm9hSZ9vpa/Q95QeOF3wkGfMUxK0tBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3d9464fe4cc8548e464df70f34d80526a0ae630b90a2a55883131b676d750d00","last_reissued_at":"2026-07-05T11:02:48.129802Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:02:48.129802Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Prioritizing Image-Related Tokens Enhances Vision-Language Pre-Training","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Hao Peng, Heng Ji, Tong Zhang, Yangyi Chen","submitted_at":"2025-05-13T21:27:52Z","abstract_excerpt":"In standard large vision-language models (LVLMs) pre-training, the model typically maximizes the joint probability of the caption conditioned on the image via next-token prediction (NTP); however, since only a small subset of caption tokens directly relates to the visual content, this naive NTP unintentionally fits the model to noise and increases the risk of hallucination. We present PRIOR, a simple vision-language pre-training approach that addresses this issue by prioritizing image-related tokens through differential weighting in the NTP loss, drawing from the importance sampling framework."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.08971","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/2505.08971/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":"2505.08971","created_at":"2026-07-05T11:02:48.129866+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.08971v1","created_at":"2026-07-05T11:02:48.129866+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.08971","created_at":"2026-07-05T11:02:48.129866+00:00"},{"alias_kind":"pith_short_12","alias_value":"HWKGJ7SMZBKI","created_at":"2026-07-05T11:02:48.129866+00:00"},{"alias_kind":"pith_short_16","alias_value":"HWKGJ7SMZBKI4RSN","created_at":"2026-07-05T11:02:48.129866+00:00"},{"alias_kind":"pith_short_8","alias_value":"HWKGJ7SM","created_at":"2026-07-05T11:02:48.129866+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/HWKGJ7SMZBKI4RSN64HTJWAFE2","json":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2.json","graph_json":"https://pith.science/api/pith-number/HWKGJ7SMZBKI4RSN64HTJWAFE2/graph.json","events_json":"https://pith.science/api/pith-number/HWKGJ7SMZBKI4RSN64HTJWAFE2/events.json","paper":"https://pith.science/paper/HWKGJ7SM"},"agent_actions":{"view_html":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2","download_json":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2.json","view_paper":"https://pith.science/paper/HWKGJ7SM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.08971&json=true","fetch_graph":"https://pith.science/api/pith-number/HWKGJ7SMZBKI4RSN64HTJWAFE2/graph.json","fetch_events":"https://pith.science/api/pith-number/HWKGJ7SMZBKI4RSN64HTJWAFE2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2/action/storage_attestation","attest_author":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2/action/author_attestation","sign_citation":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2/action/citation_signature","submit_replication":"https://pith.science/pith/HWKGJ7SMZBKI4RSN64HTJWAFE2/action/replication_record"}},"created_at":"2026-07-05T11:02:48.129866+00:00","updated_at":"2026-07-05T11:02:48.129866+00:00"}