{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Y2E7Q5ZSTW7JU7LRI5YRSJYMFS","short_pith_number":"pith:Y2E7Q5ZS","schema_version":"1.0","canonical_sha256":"c689f877329dbe9a7d71477119270c2cab52d1a4e964352851c2b40573cbda4c","source":{"kind":"arxiv","id":"2410.03450","version":2},"attestation_state":"computed","paper":{"title":"MLLM as Retriever: Interactively Learning Multimodal Retrieval for Embodied Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"B\\\"orje F. Karlsson, Junpeng Yue, Xinrun Xu, Zongqing Lu","submitted_at":"2024-10-04T14:10:39Z","abstract_excerpt":"MLLM agents demonstrate potential for complex embodied tasks by retrieving multimodal task-relevant trajectory data. However, current retrieval methods primarily focus on surface-level similarities of textual or visual cues in trajectories, neglecting their effectiveness for the specific task at hand. To address this issue, we propose a novel method, MLLM As ReTriever (MART), which enhances the performance of embodied agents by utilizing interaction data to fine-tune an MLLM retriever based on preference learning, such that the retriever fully considers the effectiveness of trajectories and pr"},"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":"2410.03450","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-04T14:10:39Z","cross_cats_sorted":[],"title_canon_sha256":"e6d13dc120d9f8a5357bfcad4f88d49fd77ab5657fda47ff1ff16ebf470b464d","abstract_canon_sha256":"97742730f90199c2f92301a371ee66f377b24448c8d4eb495d8a0d0a25d1dbc2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:07:04.948633Z","signature_b64":"U2DMaPCNmed9qNkp5pBhWov7gKoDv7ZW+zHInKpNB+pnWm96g+PQ4uuF/pl37R8hsFVol44QI5pLxnM+XTINAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c689f877329dbe9a7d71477119270c2cab52d1a4e964352851c2b40573cbda4c","last_reissued_at":"2026-07-05T11:07:04.948150Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:07:04.948150Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MLLM as Retriever: Interactively Learning Multimodal Retrieval for Embodied Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"B\\\"orje F. Karlsson, Junpeng Yue, Xinrun Xu, Zongqing Lu","submitted_at":"2024-10-04T14:10:39Z","abstract_excerpt":"MLLM agents demonstrate potential for complex embodied tasks by retrieving multimodal task-relevant trajectory data. However, current retrieval methods primarily focus on surface-level similarities of textual or visual cues in trajectories, neglecting their effectiveness for the specific task at hand. To address this issue, we propose a novel method, MLLM As ReTriever (MART), which enhances the performance of embodied agents by utilizing interaction data to fine-tune an MLLM retriever based on preference learning, such that the retriever fully considers the effectiveness of trajectories and pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03450","kind":"arxiv","version":2},"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/2410.03450/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":"2410.03450","created_at":"2026-07-05T11:07:04.948210+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.03450v2","created_at":"2026-07-05T11:07:04.948210+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03450","created_at":"2026-07-05T11:07:04.948210+00:00"},{"alias_kind":"pith_short_12","alias_value":"Y2E7Q5ZSTW7J","created_at":"2026-07-05T11:07:04.948210+00:00"},{"alias_kind":"pith_short_16","alias_value":"Y2E7Q5ZSTW7JU7LR","created_at":"2026-07-05T11:07:04.948210+00:00"},{"alias_kind":"pith_short_8","alias_value":"Y2E7Q5ZS","created_at":"2026-07-05T11:07:04.948210+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.20072","citing_title":"Extract Information from Hybrid Long Documents Leveraging LLMs: A Framework and Dataset","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS","json":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS.json","graph_json":"https://pith.science/api/pith-number/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/graph.json","events_json":"https://pith.science/api/pith-number/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/events.json","paper":"https://pith.science/paper/Y2E7Q5ZS"},"agent_actions":{"view_html":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS","download_json":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS.json","view_paper":"https://pith.science/paper/Y2E7Q5ZS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.03450&json=true","fetch_graph":"https://pith.science/api/pith-number/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/graph.json","fetch_events":"https://pith.science/api/pith-number/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/action/storage_attestation","attest_author":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/action/author_attestation","sign_citation":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/action/citation_signature","submit_replication":"https://pith.science/pith/Y2E7Q5ZSTW7JU7LRI5YRSJYMFS/action/replication_record"}},"created_at":"2026-07-05T11:07:04.948210+00:00","updated_at":"2026-07-05T11:07:04.948210+00:00"}