{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:563GZFCFEZKCIJURH2NJHMT5VB","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"e0bd3abf0cac252012b1e5dedbc45c81ef6000de8b84fd82cd03d4cfa3929448","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-02T15:01:55Z","title_canon_sha256":"a377eff375c1d10ef66388c47033dc7b9cd697db0aff551e9a1c4a83cbc1c44d"},"schema_version":"1.0","source":{"id":"2407.02333","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.02333","created_at":"2026-07-05T08:39:22Z"},{"alias_kind":"arxiv_version","alias_value":"2407.02333v1","created_at":"2026-07-05T08:39:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.02333","created_at":"2026-07-05T08:39:22Z"},{"alias_kind":"pith_short_12","alias_value":"563GZFCFEZKC","created_at":"2026-07-05T08:39:22Z"},{"alias_kind":"pith_short_16","alias_value":"563GZFCFEZKCIJUR","created_at":"2026-07-05T08:39:22Z"},{"alias_kind":"pith_short_8","alias_value":"563GZFCF","created_at":"2026-07-05T08:39:22Z"}],"graph_snapshots":[{"event_id":"sha256:dbe35cb96714c95d297e3c926c927a2be37eea4170770f3f19ba1714dd81721f","target":"graph","created_at":"2026-07-05T08:39:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2407.02333/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We uncover a surprising multilingual bias occurring in a popular class of multimodal vision-language models (VLMs). Including an image in the query to a LLaVA-style VLM significantly increases the likelihood of the model returning an English response, regardless of the language of the query. This paper investigates the causes of this loss with a two-pronged approach that combines extensive ablation of the design space with a mechanistic analysis of the models' internal representations of image and text inputs. Both approaches indicate that the issue stems in the language modelling component of","authors_text":"Anahita Bhiwandiwalla, Anne Lauscher, Carolin Holtermann, Florian Schneider, Matthew Lyle Olson, Musashi Hinck, Shaoyen Tseng, Sungduk Yu, Vasudev Lal","cross_cats":["cs.CV"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-02T15:01:55Z","title":"Why do LLaVA Vision-Language Models Reply to Images in English?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.02333","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:01a9b03ffad10754010c866c71657ee113189644a384045b35d18afbbae4aeb2","target":"record","created_at":"2026-07-05T08:39:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"e0bd3abf0cac252012b1e5dedbc45c81ef6000de8b84fd82cd03d4cfa3929448","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-02T15:01:55Z","title_canon_sha256":"a377eff375c1d10ef66388c47033dc7b9cd697db0aff551e9a1c4a83cbc1c44d"},"schema_version":"1.0","source":{"id":"2407.02333","kind":"arxiv","version":1}},"canonical_sha256":"efb66c944526542426913e9a93b27da87a3698fcdc78b91d37d59c62dbcdfb6b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"efb66c944526542426913e9a93b27da87a3698fcdc78b91d37d59c62dbcdfb6b","first_computed_at":"2026-07-05T08:39:22.830094Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:39:22.830094Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0+GCHEsNQyrRMYxua4syYrpFho4CrDJh5UrIBKCwbzbSIpLYOvUdvZ5fdDIVOYnwhjef9NMNLV8taAtRBSwqCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:39:22.830598Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.02333","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:01a9b03ffad10754010c866c71657ee113189644a384045b35d18afbbae4aeb2","sha256:dbe35cb96714c95d297e3c926c927a2be37eea4170770f3f19ba1714dd81721f"],"state_sha256":"ab155e7a2f02cd72657a2b521bd16e03debf37d10cbeac718df08046988c87a4"}