{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:BSEDKFO4REU5MIFC3FXZEP2GJK","short_pith_number":"pith:BSEDKFO4","schema_version":"1.0","canonical_sha256":"0c883515dc8929d620a2d96f923f464a808cfa519a63059c11b8ab19b465b2ea","source":{"kind":"arxiv","id":"2406.05821","version":3},"attestation_state":"computed","paper":{"title":"F-LMM: Grounding Frozen Large Multimodal Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Change Loy, Lumin Xu, Sheng Jin, Size Wu, Wei Li, Wentao Liu, Wenwei Zhang","submitted_at":"2024-06-09T15:14:26Z","abstract_excerpt":"Endowing Large Multimodal Models (LMMs) with visual grounding capability can significantly enhance AIs' understanding of the visual world and their interaction with humans. However, existing methods typically fine-tune the parameters of LMMs to learn additional segmentation tokens and overfit grounding and segmentation datasets. Such a design would inevitably cause a catastrophic diminution in the indispensable conversational capability of general AI assistants. In this paper, we comprehensively evaluate state-of-the-art grounding LMMs across a suite of multimodal question-answering benchmarks"},"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":"2406.05821","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-09T15:14:26Z","cross_cats_sorted":[],"title_canon_sha256":"1a9fbd6051f35a964c21150d15c20e1b6cd4a05c939174e93c2590564db513d2","abstract_canon_sha256":"621db11bc7042bc15b0cffdd9d00c03e45631dd9e26568cac2529deeddf946e1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:47:29.286866Z","signature_b64":"kQ60nnTaqRnorOrqfwDklpM4w0rlztMcbAMMdzNtpycz96WmI7TOGx5y0tXC0PBlZUyAmbdMn+jzpmqEbj+5BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0c883515dc8929d620a2d96f923f464a808cfa519a63059c11b8ab19b465b2ea","last_reissued_at":"2026-07-05T10:47:29.286362Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:47:29.286362Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"F-LMM: Grounding Frozen Large Multimodal Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Change Loy, Lumin Xu, Sheng Jin, Size Wu, Wei Li, Wentao Liu, Wenwei Zhang","submitted_at":"2024-06-09T15:14:26Z","abstract_excerpt":"Endowing Large Multimodal Models (LMMs) with visual grounding capability can significantly enhance AIs' understanding of the visual world and their interaction with humans. However, existing methods typically fine-tune the parameters of LMMs to learn additional segmentation tokens and overfit grounding and segmentation datasets. Such a design would inevitably cause a catastrophic diminution in the indispensable conversational capability of general AI assistants. In this paper, we comprehensively evaluate state-of-the-art grounding LMMs across a suite of multimodal question-answering benchmarks"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.05821","kind":"arxiv","version":3},"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/2406.05821/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":"2406.05821","created_at":"2026-07-05T10:47:29.286426+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.05821v3","created_at":"2026-07-05T10:47:29.286426+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.05821","created_at":"2026-07-05T10:47:29.286426+00:00"},{"alias_kind":"pith_short_12","alias_value":"BSEDKFO4REU5","created_at":"2026-07-05T10:47:29.286426+00:00"},{"alias_kind":"pith_short_16","alias_value":"BSEDKFO4REU5MIFC","created_at":"2026-07-05T10:47:29.286426+00:00"},{"alias_kind":"pith_short_8","alias_value":"BSEDKFO4","created_at":"2026-07-05T10:47:29.286426+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.20122","citing_title":"MEBench: A Novel Benchmark for Understanding Mutual Exclusivity Bias in Vision-Language Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2501.04001","citing_title":"Sa2VA: Marrying SAM2 with LLaVA for Dense Grounded Understanding of Images and Videos","ref_index":95,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK","json":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK.json","graph_json":"https://pith.science/api/pith-number/BSEDKFO4REU5MIFC3FXZEP2GJK/graph.json","events_json":"https://pith.science/api/pith-number/BSEDKFO4REU5MIFC3FXZEP2GJK/events.json","paper":"https://pith.science/paper/BSEDKFO4"},"agent_actions":{"view_html":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK","download_json":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK.json","view_paper":"https://pith.science/paper/BSEDKFO4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.05821&json=true","fetch_graph":"https://pith.science/api/pith-number/BSEDKFO4REU5MIFC3FXZEP2GJK/graph.json","fetch_events":"https://pith.science/api/pith-number/BSEDKFO4REU5MIFC3FXZEP2GJK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK/action/storage_attestation","attest_author":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK/action/author_attestation","sign_citation":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK/action/citation_signature","submit_replication":"https://pith.science/pith/BSEDKFO4REU5MIFC3FXZEP2GJK/action/replication_record"}},"created_at":"2026-07-05T10:47:29.286426+00:00","updated_at":"2026-07-05T10:47:29.286426+00:00"}