{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5732KLMANDJXBCU5DFAN2Z6BWK","short_pith_number":"pith:5732KLMA","schema_version":"1.0","canonical_sha256":"eff7a52d8068d3708a9d1940dd67c1b28553ffd6c2478ee022a0269bc5b87a97","source":{"kind":"arxiv","id":"2507.02994","version":1},"attestation_state":"computed","paper":{"title":"MedGround-R1: Advancing Medical Image Grounding via Spatial-Semantic Rewarded Group Relative Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Hongqiu Wang, Hualiang Wang, Huihui Xu, Jiyao Liu, Junjun He, Junzhi Ning, Lei Zhu, Lihao Liu, Wei Li, Xiaomeng Li, Ying Chen, Yuanpeng Nie","submitted_at":"2025-07-01T21:51:42Z","abstract_excerpt":"Medical Image Grounding (MIG), which involves localizing specific regions in medical images based on textual descriptions, requires models to not only perceive regions but also deduce spatial relationships of these regions. Existing Vision-Language Models (VLMs) for MIG often rely on Supervised Fine-Tuning (SFT) with large amounts of Chain-of-Thought (CoT) reasoning annotations, which are expensive and time-consuming to acquire. Recently, DeepSeek-R1 demonstrated that Large Language Models (LLMs) can acquire reasoning abilities through Group Relative Policy Optimization (GRPO) without requirin"},"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":"2507.02994","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-01T21:51:42Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"7fe7899bd7b52899ab5995cdf1da548d5c0fbfa757e7aa091690c1e4e3398dc7","abstract_canon_sha256":"7bca143c3bea535bb7470068ae82af9327a6f1d36c64c35c6d872416fadc059b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:31:50.416024Z","signature_b64":"CA9FibW4ZtrdJsGsp46P/EtQMQU7P0+X/8oXoibMSf/sgMyzUOluekTPHfSxEfag9G7PQZTI7udtCmbkax4SAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eff7a52d8068d3708a9d1940dd67c1b28553ffd6c2478ee022a0269bc5b87a97","last_reissued_at":"2026-07-05T11:31:50.415530Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:31:50.415530Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MedGround-R1: Advancing Medical Image Grounding via Spatial-Semantic Rewarded Group Relative Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Hongqiu Wang, Hualiang Wang, Huihui Xu, Jiyao Liu, Junjun He, Junzhi Ning, Lei Zhu, Lihao Liu, Wei Li, Xiaomeng Li, Ying Chen, Yuanpeng Nie","submitted_at":"2025-07-01T21:51:42Z","abstract_excerpt":"Medical Image Grounding (MIG), which involves localizing specific regions in medical images based on textual descriptions, requires models to not only perceive regions but also deduce spatial relationships of these regions. Existing Vision-Language Models (VLMs) for MIG often rely on Supervised Fine-Tuning (SFT) with large amounts of Chain-of-Thought (CoT) reasoning annotations, which are expensive and time-consuming to acquire. Recently, DeepSeek-R1 demonstrated that Large Language Models (LLMs) can acquire reasoning abilities through Group Relative Policy Optimization (GRPO) without requirin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.02994","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/2507.02994/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":"2507.02994","created_at":"2026-07-05T11:31:50.415589+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.02994v1","created_at":"2026-07-05T11:31:50.415589+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.02994","created_at":"2026-07-05T11:31:50.415589+00:00"},{"alias_kind":"pith_short_12","alias_value":"5732KLMANDJX","created_at":"2026-07-05T11:31:50.415589+00:00"},{"alias_kind":"pith_short_16","alias_value":"5732KLMANDJXBCU5","created_at":"2026-07-05T11:31:50.415589+00:00"},{"alias_kind":"pith_short_8","alias_value":"5732KLMA","created_at":"2026-07-05T11:31:50.415589+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.15951","citing_title":"From Failure to Feedback: Group Revision Unlocks Hard Cases in Object-Level Grounding","ref_index":83,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK","json":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK.json","graph_json":"https://pith.science/api/pith-number/5732KLMANDJXBCU5DFAN2Z6BWK/graph.json","events_json":"https://pith.science/api/pith-number/5732KLMANDJXBCU5DFAN2Z6BWK/events.json","paper":"https://pith.science/paper/5732KLMA"},"agent_actions":{"view_html":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK","download_json":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK.json","view_paper":"https://pith.science/paper/5732KLMA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.02994&json=true","fetch_graph":"https://pith.science/api/pith-number/5732KLMANDJXBCU5DFAN2Z6BWK/graph.json","fetch_events":"https://pith.science/api/pith-number/5732KLMANDJXBCU5DFAN2Z6BWK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK/action/storage_attestation","attest_author":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK/action/author_attestation","sign_citation":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK/action/citation_signature","submit_replication":"https://pith.science/pith/5732KLMANDJXBCU5DFAN2Z6BWK/action/replication_record"}},"created_at":"2026-07-05T11:31:50.415589+00:00","updated_at":"2026-07-05T11:31:50.415589+00:00"}