{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:54FYWGP3HXZ4GMNYV3AIQ5BYUU","short_pith_number":"pith:54FYWGP3","schema_version":"1.0","canonical_sha256":"ef0b8b19fb3df3c331b8aec0887438a519259bcda75a58d4f661f2822f972842","source":{"kind":"arxiv","id":"2606.01612","version":1},"attestation_state":"computed","paper":{"title":"Self-Improving Small Object Grounding in LVLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jin Sun, Ninghao Liu, Ruitong Sun, Tianze Yang, Yucheng Shi","submitted_at":"2026-06-01T03:01:38Z","abstract_excerpt":"Can internal attention patterns in Large Vision Language Models (LVLMs) identify reliable small-object boxes without fine-tuning? In this work, we provide an affirmative answer. Attention structure in LVLMs encodes grounding quality-a lightweight IoU regressor trained solely on attention maps achieves strong IoU prediction (Pearson r > 0.67). This regressor powers the regressor-based variant of our Attention-based Candidate Selection (ACS) framework, called ACS-Learned, which selects the best box from multiple sampled candidates to improve object grounding. By analyzing what the regressor lear"},"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":"2606.01612","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-01T03:01:38Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"33c3e76196fc37dfb96315dee59c9f69dfb9b819499e87a6b7f9c39f9b03962e","abstract_canon_sha256":"f28ebeed468f713d76d8120a26c6ca6a22eccdeed568d402d5703ab4b6f57df5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:38.026212Z","signature_b64":"Dh13gyzkHEtQs2yRI9xLRPxr83zwm03274h/jvOFzCPRbU7lppOZzRqFFdU8jPFhvndglqpL3jUAlPN0QbctDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ef0b8b19fb3df3c331b8aec0887438a519259bcda75a58d4f661f2822f972842","last_reissued_at":"2026-06-02T02:04:38.025803Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:38.025803Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Improving Small Object Grounding in LVLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jin Sun, Ninghao Liu, Ruitong Sun, Tianze Yang, Yucheng Shi","submitted_at":"2026-06-01T03:01:38Z","abstract_excerpt":"Can internal attention patterns in Large Vision Language Models (LVLMs) identify reliable small-object boxes without fine-tuning? In this work, we provide an affirmative answer. Attention structure in LVLMs encodes grounding quality-a lightweight IoU regressor trained solely on attention maps achieves strong IoU prediction (Pearson r > 0.67). This regressor powers the regressor-based variant of our Attention-based Candidate Selection (ACS) framework, called ACS-Learned, which selects the best box from multiple sampled candidates to improve object grounding. By analyzing what the regressor lear"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01612","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/2606.01612/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":"2606.01612","created_at":"2026-06-02T02:04:38.025862+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01612v1","created_at":"2026-06-02T02:04:38.025862+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01612","created_at":"2026-06-02T02:04:38.025862+00:00"},{"alias_kind":"pith_short_12","alias_value":"54FYWGP3HXZ4","created_at":"2026-06-02T02:04:38.025862+00:00"},{"alias_kind":"pith_short_16","alias_value":"54FYWGP3HXZ4GMNY","created_at":"2026-06-02T02:04:38.025862+00:00"},{"alias_kind":"pith_short_8","alias_value":"54FYWGP3","created_at":"2026-06-02T02:04:38.025862+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/54FYWGP3HXZ4GMNYV3AIQ5BYUU","json":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU.json","graph_json":"https://pith.science/api/pith-number/54FYWGP3HXZ4GMNYV3AIQ5BYUU/graph.json","events_json":"https://pith.science/api/pith-number/54FYWGP3HXZ4GMNYV3AIQ5BYUU/events.json","paper":"https://pith.science/paper/54FYWGP3"},"agent_actions":{"view_html":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU","download_json":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU.json","view_paper":"https://pith.science/paper/54FYWGP3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01612&json=true","fetch_graph":"https://pith.science/api/pith-number/54FYWGP3HXZ4GMNYV3AIQ5BYUU/graph.json","fetch_events":"https://pith.science/api/pith-number/54FYWGP3HXZ4GMNYV3AIQ5BYUU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU/action/storage_attestation","attest_author":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU/action/author_attestation","sign_citation":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU/action/citation_signature","submit_replication":"https://pith.science/pith/54FYWGP3HXZ4GMNYV3AIQ5BYUU/action/replication_record"}},"created_at":"2026-06-02T02:04:38.025862+00:00","updated_at":"2026-06-02T02:04:38.025862+00:00"}