{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:O3SRUOXL5F33HCXO23TIIM6P5M","short_pith_number":"pith:O3SRUOXL","schema_version":"1.0","canonical_sha256":"76e51a3aebe977b38aeed6e68433cfeb32dcd5793d19430e3c01afed0a616d96","source":{"kind":"arxiv","id":"2606.06037","version":1},"attestation_state":"computed","paper":{"title":"SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"David Ifeoluwa Adelani, Virginia Ceccatelli, Yejin Jeon","submitted_at":"2026-06-04T11:31:38Z","abstract_excerpt":"Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily evaluated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking across multiple state-of-the-art LALMs. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted aroun"},"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":"2606.06037","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2026-06-04T11:31:38Z","cross_cats_sorted":["eess.AS"],"title_canon_sha256":"38cbb549030493a647013a32da21ebdfd8a78b2c6c00cf5e45e545d4070f79d1","abstract_canon_sha256":"57cd58eea9a6addf77a9ded23899eec805a405d07b62ce6b47e243e4fcbe6b8d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:15:31.172119Z","signature_b64":"VMcAKQHZjBU8UhOw9T10ijimCRF1VsORsvlJbltP3cY6vJDXyqlOr64d54+5sxofjsuwULu+QledfuACZlYHAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76e51a3aebe977b38aeed6e68433cfeb32dcd5793d19430e3c01afed0a616d96","last_reissued_at":"2026-06-05T01:15:31.171722Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:15:31.171722Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SpeechJBB: Probing Safety Alignment and Comprehension in Large Audio Language Models under Code-Switched Speech","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"David Ifeoluwa Adelani, Virginia Ceccatelli, Yejin Jeon","submitted_at":"2026-06-04T11:31:38Z","abstract_excerpt":"Large audio language models (LALMs) are increasingly deployed in real-world applications, yet their safety alignment is still primarily evaluated on monolingual, text-based harmful prompts. This leaves their generalizability under multilingual and spoken settings, particularly code-switched speech, largely underexplored. To address this gap, we introduce SpeechJBB, an audio jailbreak dataset for benchmarking across multiple state-of-the-art LALMs. The extent of safety weaknesses is further probed by introducing an augmented setting where phonologically plausible pseudo-words are inserted aroun"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06037","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/2606.06037/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":"2606.06037","created_at":"2026-06-05T01:15:31.171787+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.06037v1","created_at":"2026-06-05T01:15:31.171787+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06037","created_at":"2026-06-05T01:15:31.171787+00:00"},{"alias_kind":"pith_short_12","alias_value":"O3SRUOXL5F33","created_at":"2026-06-05T01:15:31.171787+00:00"},{"alias_kind":"pith_short_16","alias_value":"O3SRUOXL5F33HCXO","created_at":"2026-06-05T01:15:31.171787+00:00"},{"alias_kind":"pith_short_8","alias_value":"O3SRUOXL","created_at":"2026-06-05T01:15:31.171787+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/O3SRUOXL5F33HCXO23TIIM6P5M","json":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M.json","graph_json":"https://pith.science/api/pith-number/O3SRUOXL5F33HCXO23TIIM6P5M/graph.json","events_json":"https://pith.science/api/pith-number/O3SRUOXL5F33HCXO23TIIM6P5M/events.json","paper":"https://pith.science/paper/O3SRUOXL"},"agent_actions":{"view_html":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M","download_json":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M.json","view_paper":"https://pith.science/paper/O3SRUOXL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.06037&json=true","fetch_graph":"https://pith.science/api/pith-number/O3SRUOXL5F33HCXO23TIIM6P5M/graph.json","fetch_events":"https://pith.science/api/pith-number/O3SRUOXL5F33HCXO23TIIM6P5M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M/action/storage_attestation","attest_author":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M/action/author_attestation","sign_citation":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M/action/citation_signature","submit_replication":"https://pith.science/pith/O3SRUOXL5F33HCXO23TIIM6P5M/action/replication_record"}},"created_at":"2026-06-05T01:15:31.171787+00:00","updated_at":"2026-06-05T01:15:31.171787+00:00"}