{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DTKEYHOSQSRYVHVRLQCIJ6BSYP","short_pith_number":"pith:DTKEYHOS","schema_version":"1.0","canonical_sha256":"1cd44c1dd284a38a9eb15c0484f832c3e4be98aa1c3f636ae767d2eb1b1cbbaa","source":{"kind":"arxiv","id":"2502.11596","version":1},"attestation_state":"computed","paper":{"title":"LLM Embeddings for Deep Learning on Tabular Data","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andrei Margeloiu, Bla\\v{z} \\v{S}krlj, Boshko Koloski, Mateja Jamnik, Nikola Simidjievski, Xiangjian Jiang","submitted_at":"2025-02-17T09:28:51Z","abstract_excerpt":"Tabular deep-learning methods require embedding numerical and categorical input features into high-dimensional spaces before processing them. Existing methods deal with this heterogeneous nature of tabular data by employing separate type-specific encoding approaches. This limits the cross-table transfer potential and the exploitation of pre-trained knowledge. We propose a novel approach that first transforms tabular data into text, and then leverages pre-trained representations from LLMs to encode this data, resulting in a plug-and-play solution to improv ing deep-learning tabular methods. We "},"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":"2502.11596","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T09:28:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"3e7043b142928e0ad91b85b4bdd4cb97ba7a563062ce0fa3cab2bb7887774ff9","abstract_canon_sha256":"9368ba28019a2ad4164e00da913bbdf0b02a09cce53a5fe16474123046582e19"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:31.381944Z","signature_b64":"cWWm3xb8hqbsSnmKAuh1Fqm9haOQC8XFIDR44iHVqRdqavFQes1DYU7gCCXo5aJwLIlL8mc9/zb+wKZ3gGhJDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1cd44c1dd284a38a9eb15c0484f832c3e4be98aa1c3f636ae767d2eb1b1cbbaa","last_reissued_at":"2026-07-05T10:15:31.381400Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:31.381400Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM Embeddings for Deep Learning on Tabular Data","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andrei Margeloiu, Bla\\v{z} \\v{S}krlj, Boshko Koloski, Mateja Jamnik, Nikola Simidjievski, Xiangjian Jiang","submitted_at":"2025-02-17T09:28:51Z","abstract_excerpt":"Tabular deep-learning methods require embedding numerical and categorical input features into high-dimensional spaces before processing them. Existing methods deal with this heterogeneous nature of tabular data by employing separate type-specific encoding approaches. This limits the cross-table transfer potential and the exploitation of pre-trained knowledge. We propose a novel approach that first transforms tabular data into text, and then leverages pre-trained representations from LLMs to encode this data, resulting in a plug-and-play solution to improv ing deep-learning tabular methods. We "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11596","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/2502.11596/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":"2502.11596","created_at":"2026-07-05T10:15:31.381454+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.11596v1","created_at":"2026-07-05T10:15:31.381454+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11596","created_at":"2026-07-05T10:15:31.381454+00:00"},{"alias_kind":"pith_short_12","alias_value":"DTKEYHOSQSRY","created_at":"2026-07-05T10:15:31.381454+00:00"},{"alias_kind":"pith_short_16","alias_value":"DTKEYHOSQSRYVHVR","created_at":"2026-07-05T10:15:31.381454+00:00"},{"alias_kind":"pith_short_8","alias_value":"DTKEYHOS","created_at":"2026-07-05T10:15:31.381454+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12411","citing_title":"Predicting Decisions of AI Agents from Limited Interaction through Text-Tabular Modeling","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11835","citing_title":"Schema-Adaptive Tabular Representation Learning with LLMs for Generalizable Multimodal Clinical Reasoning","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP","json":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP.json","graph_json":"https://pith.science/api/pith-number/DTKEYHOSQSRYVHVRLQCIJ6BSYP/graph.json","events_json":"https://pith.science/api/pith-number/DTKEYHOSQSRYVHVRLQCIJ6BSYP/events.json","paper":"https://pith.science/paper/DTKEYHOS"},"agent_actions":{"view_html":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP","download_json":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP.json","view_paper":"https://pith.science/paper/DTKEYHOS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.11596&json=true","fetch_graph":"https://pith.science/api/pith-number/DTKEYHOSQSRYVHVRLQCIJ6BSYP/graph.json","fetch_events":"https://pith.science/api/pith-number/DTKEYHOSQSRYVHVRLQCIJ6BSYP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP/action/storage_attestation","attest_author":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP/action/author_attestation","sign_citation":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP/action/citation_signature","submit_replication":"https://pith.science/pith/DTKEYHOSQSRYVHVRLQCIJ6BSYP/action/replication_record"}},"created_at":"2026-07-05T10:15:31.381454+00:00","updated_at":"2026-07-05T10:15:31.381454+00:00"}