{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PTO3L2DE7FLWFJWHVBIRL2QDCJ","short_pith_number":"pith:PTO3L2DE","schema_version":"1.0","canonical_sha256":"7cddb5e864f95762a6c7a85115ea031264748a0368fb3f9e2960e5c690b83367","source":{"kind":"arxiv","id":"2607.18988","version":1},"attestation_state":"computed","paper":{"title":"DobicVLM: Aligning Chest X-Ray Report Generation with Clinically-Grounded Programmatic Rewards via Group Relative Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Abiodun Adereni, Ademide Animasaun, Andem Aniekan, Angel Ezendu, Angelica Obayi, Christian Maurice, Ephraim Usani, Mary Donald Essien, Oluwasegun Oguntuase, Oluwaseun Odunsi, Philip Chibundu, Samuel Okoko, Thanni Adewuyi","submitted_at":"2026-07-21T11:20:09Z","abstract_excerpt":"Medical imaging is a cornerstone of diagnostics, yet automated chest X-ray report generation struggles with structural adherence, anatomical completeness, and semantic faithfulness. We introduce DobicVLM, a vision-language model combining supervised fine-tuning on MedGemma-4B with Group Relative Policy Optimization (GRPO) and clinically-grounded programmatic rewards. Our approach uses interpretable, rule-based reward components; structural verification, anatomical checklist, semantic similarity, and length constraints to enforce clinical standards without neural reward models. Trained on 1,000"},"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":"2607.18988","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-21T11:20:09Z","cross_cats_sorted":[],"title_canon_sha256":"6ad3d8a6a88cb270b649c955df80c6768325d85af133a4de0a1be5112ccf4e19","abstract_canon_sha256":"36b301f7318cd8724eb701f2d2d41299b3da569120ab88f891aca540a715fa93"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-22T01:24:03.179822Z","signature_b64":"VvQ2oTy9/R4P4hemRSXdZ/1BNriZXFE/UybYmypBxAGNbFYxW2BihK3N7sDGQ8i/lYfumiCle9WAdoe8gbuTAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7cddb5e864f95762a6c7a85115ea031264748a0368fb3f9e2960e5c690b83367","last_reissued_at":"2026-07-22T01:24:03.178978Z","signature_status":"signed_v1","first_computed_at":"2026-07-22T01:24:03.178978Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DobicVLM: Aligning Chest X-Ray Report Generation with Clinically-Grounded Programmatic Rewards via Group Relative Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Abiodun Adereni, Ademide Animasaun, Andem Aniekan, Angel Ezendu, Angelica Obayi, Christian Maurice, Ephraim Usani, Mary Donald Essien, Oluwasegun Oguntuase, Oluwaseun Odunsi, Philip Chibundu, Samuel Okoko, Thanni Adewuyi","submitted_at":"2026-07-21T11:20:09Z","abstract_excerpt":"Medical imaging is a cornerstone of diagnostics, yet automated chest X-ray report generation struggles with structural adherence, anatomical completeness, and semantic faithfulness. We introduce DobicVLM, a vision-language model combining supervised fine-tuning on MedGemma-4B with Group Relative Policy Optimization (GRPO) and clinically-grounded programmatic rewards. Our approach uses interpretable, rule-based reward components; structural verification, anatomical checklist, semantic similarity, and length constraints to enforce clinical standards without neural reward models. Trained on 1,000"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.18988","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/2607.18988/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":"2607.18988","created_at":"2026-07-22T01:24:03.179384+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.18988v1","created_at":"2026-07-22T01:24:03.179384+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.18988","created_at":"2026-07-22T01:24:03.179384+00:00"},{"alias_kind":"pith_short_12","alias_value":"PTO3L2DE7FLW","created_at":"2026-07-22T01:24:03.179384+00:00"},{"alias_kind":"pith_short_16","alias_value":"PTO3L2DE7FLWFJWH","created_at":"2026-07-22T01:24:03.179384+00:00"},{"alias_kind":"pith_short_8","alias_value":"PTO3L2DE","created_at":"2026-07-22T01:24:03.179384+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/PTO3L2DE7FLWFJWHVBIRL2QDCJ","json":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ.json","graph_json":"https://pith.science/api/pith-number/PTO3L2DE7FLWFJWHVBIRL2QDCJ/graph.json","events_json":"https://pith.science/api/pith-number/PTO3L2DE7FLWFJWHVBIRL2QDCJ/events.json","paper":"https://pith.science/paper/PTO3L2DE"},"agent_actions":{"view_html":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ","download_json":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ.json","view_paper":"https://pith.science/paper/PTO3L2DE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.18988&json=true","fetch_graph":"https://pith.science/api/pith-number/PTO3L2DE7FLWFJWHVBIRL2QDCJ/graph.json","fetch_events":"https://pith.science/api/pith-number/PTO3L2DE7FLWFJWHVBIRL2QDCJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ/action/storage_attestation","attest_author":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ/action/author_attestation","sign_citation":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ/action/citation_signature","submit_replication":"https://pith.science/pith/PTO3L2DE7FLWFJWHVBIRL2QDCJ/action/replication_record"}},"created_at":"2026-07-22T01:24:03.179384+00:00","updated_at":"2026-07-22T01:24:03.179384+00:00"}