{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:WKP2IC7SI5DKRF7MFITCAKBSXZ","short_pith_number":"pith:WKP2IC7S","schema_version":"1.0","canonical_sha256":"b29fa40bf24746a897ec2a26202832be47b3717917d367eab40efbd638ff0381","source":{"kind":"arxiv","id":"2605.18160","version":1},"attestation_state":"computed","paper":{"title":"Vision Inference Former: Sustaining Visual Consistency in Multimodal Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Fei Wu, Kairong Han, Kun Kuang, Min Zhang, Xinpeng Dong, Xu Tan","submitted_at":"2026-05-18T10:04:22Z","abstract_excerpt":"In recent years, multimodal large language models (MLLMs) have achieved remarkable progress, primarily attributed to effective paradigms for integrating visual and textual information. The dominant connector-based paradigm projects visual features into textual sequence, enabling unified multimodal alignment and reasoning within a generative architecture. However, our experiments reveal two key limitations: (1) Although visual information serves as the core evidential modality in MLLMs, it is treated on par with textual tokens, diminishing the unique contribution of the visual modality; (2) As "},"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":"2605.18160","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-18T10:04:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9a65c3a510b56109537f04de7d0db762e3e1cce70bccd1f0f581a9485acc1d97","abstract_canon_sha256":"1158020028430bfa009ef88169b50c1d65bf1a00a9131c7f4efdd94376567e36"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:05:48.768552Z","signature_b64":"W7SoQyG9XbzRihxXs+edoPM9wDrolnC5s56ytQRjqCh8EwzeUWFxQ3SHCTqyxoUGbEHgTytXipz7tqD8y2vsBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b29fa40bf24746a897ec2a26202832be47b3717917d367eab40efbd638ff0381","last_reissued_at":"2026-05-20T00:05:48.768026Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:05:48.768026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vision Inference Former: Sustaining Visual Consistency in Multimodal Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Fei Wu, Kairong Han, Kun Kuang, Min Zhang, Xinpeng Dong, Xu Tan","submitted_at":"2026-05-18T10:04:22Z","abstract_excerpt":"In recent years, multimodal large language models (MLLMs) have achieved remarkable progress, primarily attributed to effective paradigms for integrating visual and textual information. The dominant connector-based paradigm projects visual features into textual sequence, enabling unified multimodal alignment and reasoning within a generative architecture. However, our experiments reveal two key limitations: (1) Although visual information serves as the core evidential modality in MLLMs, it is treated on par with textual tokens, diminishing the unique contribution of the visual modality; (2) As "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18160","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/2605.18160/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-19T23:41:59.073399Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T23:33:35.362506Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"f658bd97f569116feb518ff42934a5cbd14f9c8f6ee9c64f728ffbe9bb3b5c40"},"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":"2605.18160","created_at":"2026-05-20T00:05:48.768096+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.18160v1","created_at":"2026-05-20T00:05:48.768096+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18160","created_at":"2026-05-20T00:05:48.768096+00:00"},{"alias_kind":"pith_short_12","alias_value":"WKP2IC7SI5DK","created_at":"2026-05-20T00:05:48.768096+00:00"},{"alias_kind":"pith_short_16","alias_value":"WKP2IC7SI5DKRF7M","created_at":"2026-05-20T00:05:48.768096+00:00"},{"alias_kind":"pith_short_8","alias_value":"WKP2IC7S","created_at":"2026-05-20T00:05:48.768096+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/WKP2IC7SI5DKRF7MFITCAKBSXZ","json":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ.json","graph_json":"https://pith.science/api/pith-number/WKP2IC7SI5DKRF7MFITCAKBSXZ/graph.json","events_json":"https://pith.science/api/pith-number/WKP2IC7SI5DKRF7MFITCAKBSXZ/events.json","paper":"https://pith.science/paper/WKP2IC7S"},"agent_actions":{"view_html":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ","download_json":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ.json","view_paper":"https://pith.science/paper/WKP2IC7S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.18160&json=true","fetch_graph":"https://pith.science/api/pith-number/WKP2IC7SI5DKRF7MFITCAKBSXZ/graph.json","fetch_events":"https://pith.science/api/pith-number/WKP2IC7SI5DKRF7MFITCAKBSXZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ/action/storage_attestation","attest_author":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ/action/author_attestation","sign_citation":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ/action/citation_signature","submit_replication":"https://pith.science/pith/WKP2IC7SI5DKRF7MFITCAKBSXZ/action/replication_record"}},"created_at":"2026-05-20T00:05:48.768096+00:00","updated_at":"2026-05-20T00:05:48.768096+00:00"}