{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PHEWBQ2T5J7IF6B62PUV6W2QWZ","short_pith_number":"pith:PHEWBQ2T","schema_version":"1.0","canonical_sha256":"79c960c353ea7e82f83ed3e95f5b50b66eeca300e4054fe06ccb34309e8c2f82","source":{"kind":"arxiv","id":"2606.22864","version":1},"attestation_state":"computed","paper":{"title":"When AUC 0.998 Is Not Enough: A Candidate Evaluation Protocol for Hidden-State Probes of Indirect Prompt Injection in Multimodal Computer-Use Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","submitted_at":"2026-06-22T05:13:25Z","abstract_excerpt":"Hidden-state probing -- a linear classifier on a frozen vision-language model's internal activations -- has emerged as an attractive evaluation tool for flagging indirect prompt injection (IPI) in multimodal computer-use agents before the agent emits a corrupted action. We argue, on a single-backbone cautionary case study (Qwen2.5-VL-7B on Mind2Web, teacher-forced replay), that a high probing AUC on a clean-vs-attack split is not, on its own, evidence of malicious-content detection. Two post-hoc diagnostics -- a paired-construction scalar baseline on text-side injections, and same-step nuisanc"},"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":"2606.22864","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T05:13:25Z","cross_cats_sorted":[],"title_canon_sha256":"bc947b8e35db9ec5f6950eea4c6c6939912b3bc15acb42b40a3861987721bdc5","abstract_canon_sha256":"4df5427d75552e1202436ca193751cb761708d6d8fa54fdb6185acca4f92f718"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T03:14:02.462056Z","signature_b64":"vXg/TftK7DTNyRG/dSKckpUWPfhTby3sdeKfkgXz5rh9asOBn9pm5pVBZboZiPEH+Tlv5ifCcDgpabpmsKwEAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"79c960c353ea7e82f83ed3e95f5b50b66eeca300e4054fe06ccb34309e8c2f82","last_reissued_at":"2026-06-23T03:14:02.461595Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T03:14:02.461595Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When AUC 0.998 Is Not Enough: A Candidate Evaluation Protocol for Hidden-State Probes of Indirect Prompt Injection in Multimodal Computer-Use Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","submitted_at":"2026-06-22T05:13:25Z","abstract_excerpt":"Hidden-state probing -- a linear classifier on a frozen vision-language model's internal activations -- has emerged as an attractive evaluation tool for flagging indirect prompt injection (IPI) in multimodal computer-use agents before the agent emits a corrupted action. We argue, on a single-backbone cautionary case study (Qwen2.5-VL-7B on Mind2Web, teacher-forced replay), that a high probing AUC on a clean-vs-attack split is not, on its own, evidence of malicious-content detection. Two post-hoc diagnostics -- a paired-construction scalar baseline on text-side injections, and same-step nuisanc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22864","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/2606.22864/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":"2606.22864","created_at":"2026-06-23T03:14:02.461658+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.22864v1","created_at":"2026-06-23T03:14:02.461658+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22864","created_at":"2026-06-23T03:14:02.461658+00:00"},{"alias_kind":"pith_short_12","alias_value":"PHEWBQ2T5J7I","created_at":"2026-06-23T03:14:02.461658+00:00"},{"alias_kind":"pith_short_16","alias_value":"PHEWBQ2T5J7IF6B6","created_at":"2026-06-23T03:14:02.461658+00:00"},{"alias_kind":"pith_short_8","alias_value":"PHEWBQ2T","created_at":"2026-06-23T03:14:02.461658+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/PHEWBQ2T5J7IF6B62PUV6W2QWZ","json":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ.json","graph_json":"https://pith.science/api/pith-number/PHEWBQ2T5J7IF6B62PUV6W2QWZ/graph.json","events_json":"https://pith.science/api/pith-number/PHEWBQ2T5J7IF6B62PUV6W2QWZ/events.json","paper":"https://pith.science/paper/PHEWBQ2T"},"agent_actions":{"view_html":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ","download_json":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ.json","view_paper":"https://pith.science/paper/PHEWBQ2T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.22864&json=true","fetch_graph":"https://pith.science/api/pith-number/PHEWBQ2T5J7IF6B62PUV6W2QWZ/graph.json","fetch_events":"https://pith.science/api/pith-number/PHEWBQ2T5J7IF6B62PUV6W2QWZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ/action/storage_attestation","attest_author":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ/action/author_attestation","sign_citation":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ/action/citation_signature","submit_replication":"https://pith.science/pith/PHEWBQ2T5J7IF6B62PUV6W2QWZ/action/replication_record"}},"created_at":"2026-06-23T03:14:02.461658+00:00","updated_at":"2026-06-23T03:14:02.461658+00:00"}