{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CDPXLLXDLVKC4L4DTXEESFZFVY","short_pith_number":"pith:CDPXLLXD","schema_version":"1.0","canonical_sha256":"10df75aee35d542e2f839dc8491725ae186e4a73fabe8f8a042467bb4071a8b8","source":{"kind":"arxiv","id":"2605.08245","version":3},"attestation_state":"computed","paper":{"title":"When Language Overwrites Vision: Over-Alignment and Geometric Debiasing in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Vision-language models hallucinate because they over-align visual embeddings to text, and removing a linguistic bias subspace fixes it.","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dianbo Liu, Harshvardhan Saini, Samyak Jha, Yiming Tang","submitted_at":"2026-05-07T10:09:18Z","abstract_excerpt":"Vision-Language Models (VLMs) increasingly power high-stakes applications, from medical imaging to autonomous systems, yet they routinely hallucinate, confidently describing content not present in the input. We investigate the root causes of these failure modes with a mechanistic analysis focusing on the decoder-based VLMs. We trace these failure modes to a geometric over-alignment: to bridge the modality gap required by attention mechanisms, decoder-based VLMs over-align visual embeddings with the text manifold, injecting a statistical linguistic bias that systematically overshadows fine-grai"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.08245","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-07T10:09:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1f2d578158156bab9deba077bb72a4ceb7c2d42703adca3a0e2f2827023a0c89","abstract_canon_sha256":"dd62347be14405281fa47c7f4afb2a286d61049507361e4209e6cba3f86acabc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:41.432819Z","signature_b64":"iQVPoiolNhTLYSmmWp1y4/kfu596FVDqqs6A/Z7urh08ANtViGlsWxQuV4ouUja6FUezyoqrSDe2gaV3SlE9Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"10df75aee35d542e2f839dc8491725ae186e4a73fabe8f8a042467bb4071a8b8","last_reissued_at":"2026-05-20T00:00:41.432026Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:41.432026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When Language Overwrites Vision: Over-Alignment and Geometric Debiasing in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Vision-language models hallucinate because they over-align visual embeddings to text, and removing a linguistic bias subspace fixes it.","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dianbo Liu, Harshvardhan Saini, Samyak Jha, Yiming Tang","submitted_at":"2026-05-07T10:09:18Z","abstract_excerpt":"Vision-Language Models (VLMs) increasingly power high-stakes applications, from medical imaging to autonomous systems, yet they routinely hallucinate, confidently describing content not present in the input. We investigate the root causes of these failure modes with a mechanistic analysis focusing on the decoder-based VLMs. We trace these failure modes to a geometric over-alignment: to bridge the modality gap required by attention mechanisms, decoder-based VLMs over-align visual embeddings with the text manifold, injecting a statistical linguistic bias that systematically overshadows fine-grai"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We trace these failure modes to a geometric over-alignment: to bridge the modality gap required by attention mechanisms, decoder-based VLMs over-align visual embeddings with the text manifold, injecting a statistical linguistic bias that systematically overshadows fine-grained visual evidence.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"Linguistic bias concentrates in the top principal components of a universal, dataset-agnostic text subspace, so that explicitly projecting this subspace from visual representations removes the bias without discarding task-critical visual information.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Decoder-based VLMs hallucinate because visual embeddings are over-aligned to a text manifold; projecting out the top principal components of a universal linguistic subspace reduces this bias and improves benchmark performance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Vision-language models hallucinate because they over-align visual embeddings to text, and removing a linguistic bias subspace fixes it.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f9c88b5b806f759216d3786e2a091569f7924c8d22eb035ff0b8f41d9379ce2b"},"source":{"id":"2605.08245","kind":"arxiv","version":3},"verdict":{"id":"e93a481b-41ca-4ba1-882c-196df3dda3b8","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T17:01:42.168645Z","strongest_claim":"We trace these failure modes to a geometric over-alignment: to bridge the modality gap required by attention mechanisms, decoder-based VLMs over-align visual embeddings with the text manifold, injecting a statistical linguistic bias that systematically overshadows fine-grained visual evidence.","one_line_summary":"Decoder-based VLMs hallucinate because visual embeddings are over-aligned to a text manifold; projecting out the top principal components of a universal linguistic subspace reduces this bias and improves benchmark performance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"Linguistic bias concentrates in the top principal components of a universal, dataset-agnostic text subspace, so that explicitly projecting this subspace from visual representations removes the bias without discarding task-critical visual information.","pith_extraction_headline":"Vision-language models hallucinate because they over-align visual embeddings to text, and removing a linguistic bias subspace fixes it."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.08245/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T19:31:19.305479Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T13:06:17.537298Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9c0a38d34e46c4e2b6b3736b7aa9f3551be0c448b3582020c5ac9c3288f88ef8"},"references":{"count":23,"sample":[{"doi":"","year":null,"title":"Aishwarya Agrawal, Jiasen Lu, Stanislaw Antol, Margaret Mitchell, C","work_id":"fe2a10cc-b8d6-4236-b413-9ae3cdc9470f","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Neeraj Anand, Samyak Jha, Udbhav Bamba, and Rahul Rahaman","work_id":"131b814a-7192-4c6c-893a-b6c325135269","ref_index":2,"cited_arxiv_id":"1505.00468","is_internal_anchor":true},{"doi":"","year":null,"title":"Qwen3-VL Technical Report","work_id":"1fe243aa-e3c0-4da6-b391-4cbcfc88d5c0","ref_index":3,"cited_arxiv_id":"2511.21631","is_internal_anchor":true},{"doi":"","year":2023,"title":"David M Chan, Suzanne Petryk, Joseph E Gonzalez, Trevor Darrell, and John Canny","work_id":"7733671b-ca00-46d6-ba19-73ef7f876327","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Sparse Autoencoders Find Highly Interpretable Features in Language Models","work_id":"51960d72-c69f-4db8-8efd-e90e8b4d9524","ref_index":5,"cited_arxiv_id":"2309.08600","is_internal_anchor":true}],"resolved_work":23,"snapshot_sha256":"1129fe4c1831a55e4958aeeb2e9e24c60b5f1b4d8556dff32f89f963f76f14fd","internal_anchors":9},"formal_canon":{"evidence_count":2,"snapshot_sha256":"85181579ed5b55d53a85445141e278cf394e6cae5cb6f9c17f8612369bf0549f"},"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":"2605.08245","created_at":"2026-05-20T00:00:41.432141+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.08245v3","created_at":"2026-05-20T00:00:41.432141+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.08245","created_at":"2026-05-20T00:00:41.432141+00:00"},{"alias_kind":"pith_short_12","alias_value":"CDPXLLXDLVKC","created_at":"2026-05-20T00:00:41.432141+00:00"},{"alias_kind":"pith_short_16","alias_value":"CDPXLLXDLVKC4L4D","created_at":"2026-05-20T00:00:41.432141+00:00"},{"alias_kind":"pith_short_8","alias_value":"CDPXLLXD","created_at":"2026-05-20T00:00:41.432141+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY","json":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY.json","graph_json":"https://pith.science/api/pith-number/CDPXLLXDLVKC4L4DTXEESFZFVY/graph.json","events_json":"https://pith.science/api/pith-number/CDPXLLXDLVKC4L4DTXEESFZFVY/events.json","paper":"https://pith.science/paper/CDPXLLXD"},"agent_actions":{"view_html":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY","download_json":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY.json","view_paper":"https://pith.science/paper/CDPXLLXD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.08245&json=true","fetch_graph":"https://pith.science/api/pith-number/CDPXLLXDLVKC4L4DTXEESFZFVY/graph.json","fetch_events":"https://pith.science/api/pith-number/CDPXLLXDLVKC4L4DTXEESFZFVY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY/action/storage_attestation","attest_author":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY/action/author_attestation","sign_citation":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY/action/citation_signature","submit_replication":"https://pith.science/pith/CDPXLLXDLVKC4L4DTXEESFZFVY/action/replication_record"}},"created_at":"2026-05-20T00:00:41.432141+00:00","updated_at":"2026-05-20T00:00:41.432141+00:00"}