{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:KF4OG6VXRYTBTH6BLBTO2W3XOY","short_pith_number":"pith:KF4OG6VX","schema_version":"1.0","canonical_sha256":"5178e37ab78e26199fc15866ed5b77761e4e45885b4a5c42e78e6931fa38188a","source":{"kind":"arxiv","id":"2411.14708","version":3},"attestation_state":"computed","paper":{"title":"Understanding LLM Embeddings for Regression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Bangding Yang, Eric Tang, Xingyou Song","submitted_at":"2024-11-22T03:33:51Z","abstract_excerpt":"With the rise of large language models (LLMs) for flexibly processing information as strings, a natural application is regression, specifically by preprocessing string representations into LLM embeddings as downstream features for metric prediction. In this paper, we provide one of the first comprehensive investigations into embedding-based regression and demonstrate that LLM embeddings as features can be better for high-dimensional regression tasks than using traditional feature engineering. This regression performance can be explained in part due to LLM embeddings over numeric data inherentl"},"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":"2411.14708","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-11-22T03:33:51Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"43e563458670bb9cd51ac0cab47191096880d04f9e4c118c5c802d337fc82d9e","abstract_canon_sha256":"2d81c56ad067e15fb1dbfe2bada261cc42cd673400f21f6d2be85fd9ff371aaf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:48.225916Z","signature_b64":"worwAVKQwMQxFoDnRRnyWjLkahyL0aYGm11Rqk2vNJk54cuEvAUca9S14vrP7YvZNkszYzkvwjAnmkWXiHLBBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5178e37ab78e26199fc15866ed5b77761e4e45885b4a5c42e78e6931fa38188a","last_reissued_at":"2026-07-05T10:14:48.225498Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:48.225498Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding LLM Embeddings for Regression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Bangding Yang, Eric Tang, Xingyou Song","submitted_at":"2024-11-22T03:33:51Z","abstract_excerpt":"With the rise of large language models (LLMs) for flexibly processing information as strings, a natural application is regression, specifically by preprocessing string representations into LLM embeddings as downstream features for metric prediction. In this paper, we provide one of the first comprehensive investigations into embedding-based regression and demonstrate that LLM embeddings as features can be better for high-dimensional regression tasks than using traditional feature engineering. This regression performance can be explained in part due to LLM embeddings over numeric data inherentl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.14708","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/2411.14708/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":"2411.14708","created_at":"2026-07-05T10:14:48.225544+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.14708v3","created_at":"2026-07-05T10:14:48.225544+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.14708","created_at":"2026-07-05T10:14:48.225544+00:00"},{"alias_kind":"pith_short_12","alias_value":"KF4OG6VXRYTB","created_at":"2026-07-05T10:14:48.225544+00:00"},{"alias_kind":"pith_short_16","alias_value":"KF4OG6VXRYTBTH6B","created_at":"2026-07-05T10:14:48.225544+00:00"},{"alias_kind":"pith_short_8","alias_value":"KF4OG6VX","created_at":"2026-07-05T10:14:48.225544+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22719","citing_title":"Leakage-Aware Benchmarking of LLM Forecasting: Real-Time Nowcasts as the Decision-Time Input for Macro Factor Ranking","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29900","citing_title":"LLM-based Multimodal Personality Recognition via Facial Action Unit-Text Semantic Fusion","ref_index":122,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20740","citing_title":"Distribution-Aware Reward: Reinforcement Learning over Predictive Distributions for LLM Regression","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY","json":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY.json","graph_json":"https://pith.science/api/pith-number/KF4OG6VXRYTBTH6BLBTO2W3XOY/graph.json","events_json":"https://pith.science/api/pith-number/KF4OG6VXRYTBTH6BLBTO2W3XOY/events.json","paper":"https://pith.science/paper/KF4OG6VX"},"agent_actions":{"view_html":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY","download_json":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY.json","view_paper":"https://pith.science/paper/KF4OG6VX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.14708&json=true","fetch_graph":"https://pith.science/api/pith-number/KF4OG6VXRYTBTH6BLBTO2W3XOY/graph.json","fetch_events":"https://pith.science/api/pith-number/KF4OG6VXRYTBTH6BLBTO2W3XOY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY/action/storage_attestation","attest_author":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY/action/author_attestation","sign_citation":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY/action/citation_signature","submit_replication":"https://pith.science/pith/KF4OG6VXRYTBTH6BLBTO2W3XOY/action/replication_record"}},"created_at":"2026-07-05T10:14:48.225544+00:00","updated_at":"2026-07-05T10:14:48.225544+00:00"}