{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TWPZ323EYJUFVTN5VB4QS7RC3G","short_pith_number":"pith:TWPZ323E","schema_version":"1.0","canonical_sha256":"9d9f9deb64c2685acdbda879097e22d9960abd6a69958112ae8f90026ad7160f","source":{"kind":"arxiv","id":"2606.23487","version":1},"attestation_state":"computed","paper":{"title":"CADRE: Stable, Parameter Efficient Adaptation of Medical Vision Language Models with Bounded Forgetting and Prior Drift","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Amrita Singh, Rishabh Jha","submitted_at":"2026-06-22T15:34:32Z","abstract_excerpt":"Medical vision-language models (VLMs) such as BiomedCLIP generalize broadly, but adapting them to a clinical service is as much a safety problem as an accuracy one. Updating a deployed model for a new imaging modality can fail silently in two ways that harm patients: it can forget modalities it already handled (catastrophic forgetting), and it can drift from its trustworthy pretrained prior toward modality-specific shortcuts. We study parameter-efficient continual adaptation through these two properties rather than leaderboard accuracy, presenting CADRE: a frozen-backbone framework combining l"},"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":"2606.23487","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-22T15:34:32Z","cross_cats_sorted":[],"title_canon_sha256":"3c74436b5591ddaa2ebd21b6c7466a92e4204e3c3ef74153bca4b08b9147af53","abstract_canon_sha256":"bfbdfb94445ad6e75345de36456e73bd6b93cca8539252f315f35df90064b4c2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T03:14:21.275639Z","signature_b64":"rsieBM/7qUHTvXFDk56+VpGu3zJcOTLiDW1cC2YscV/7AVoHmPexSl1KTefuMHSJU9WAuUy//tyKDU0nPLaWDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d9f9deb64c2685acdbda879097e22d9960abd6a69958112ae8f90026ad7160f","last_reissued_at":"2026-06-23T03:14:21.275242Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T03:14:21.275242Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CADRE: Stable, Parameter Efficient Adaptation of Medical Vision Language Models with Bounded Forgetting and Prior Drift","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Amrita Singh, Rishabh Jha","submitted_at":"2026-06-22T15:34:32Z","abstract_excerpt":"Medical vision-language models (VLMs) such as BiomedCLIP generalize broadly, but adapting them to a clinical service is as much a safety problem as an accuracy one. Updating a deployed model for a new imaging modality can fail silently in two ways that harm patients: it can forget modalities it already handled (catastrophic forgetting), and it can drift from its trustworthy pretrained prior toward modality-specific shortcuts. We study parameter-efficient continual adaptation through these two properties rather than leaderboard accuracy, presenting CADRE: a frozen-backbone framework combining l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.23487","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/2606.23487/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":"2606.23487","created_at":"2026-06-23T03:14:21.275298+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.23487v1","created_at":"2026-06-23T03:14:21.275298+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.23487","created_at":"2026-06-23T03:14:21.275298+00:00"},{"alias_kind":"pith_short_12","alias_value":"TWPZ323EYJUF","created_at":"2026-06-23T03:14:21.275298+00:00"},{"alias_kind":"pith_short_16","alias_value":"TWPZ323EYJUFVTN5","created_at":"2026-06-23T03:14:21.275298+00:00"},{"alias_kind":"pith_short_8","alias_value":"TWPZ323E","created_at":"2026-06-23T03:14:21.275298+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/TWPZ323EYJUFVTN5VB4QS7RC3G","json":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G.json","graph_json":"https://pith.science/api/pith-number/TWPZ323EYJUFVTN5VB4QS7RC3G/graph.json","events_json":"https://pith.science/api/pith-number/TWPZ323EYJUFVTN5VB4QS7RC3G/events.json","paper":"https://pith.science/paper/TWPZ323E"},"agent_actions":{"view_html":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G","download_json":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G.json","view_paper":"https://pith.science/paper/TWPZ323E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.23487&json=true","fetch_graph":"https://pith.science/api/pith-number/TWPZ323EYJUFVTN5VB4QS7RC3G/graph.json","fetch_events":"https://pith.science/api/pith-number/TWPZ323EYJUFVTN5VB4QS7RC3G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G/action/storage_attestation","attest_author":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G/action/author_attestation","sign_citation":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G/action/citation_signature","submit_replication":"https://pith.science/pith/TWPZ323EYJUFVTN5VB4QS7RC3G/action/replication_record"}},"created_at":"2026-06-23T03:14:21.275298+00:00","updated_at":"2026-06-23T03:14:21.275298+00:00"}