{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EN7KHWA7LQ3VBCAFBALRCZU2CC","short_pith_number":"pith:EN7KHWA7","schema_version":"1.0","canonical_sha256":"237ea3d81f5c37508805081711669a10bbefc2de8117adda8c9aa6de1b0a422e","source":{"kind":"arxiv","id":"2310.02989","version":2},"attestation_state":"computed","paper":{"title":"xVal: A Continuous Numerical Tokenization for Scientific Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"stat.ML","authors_text":"Alberto Bietti, Bruno R\\'egaldo-Saint Blancard, Francois Lanusse, Geraud Krawezik, Kyunghyun Cho, Liam Parker, Mariel Pettee, Michael Eickenberg, Michael McCabe, Miles Cranmer, Ruben Ohana, Shirley Ho, Siavash Golkar, Tiberiu Tesileanu","submitted_at":"2023-10-04T17:26:16Z","abstract_excerpt":"Due in part to their discontinuous and discrete default encodings for numbers, Large Language Models (LLMs) have not yet been commonly used to process numerically-dense scientific datasets. Rendering datasets as text, however, could help aggregate diverse and multi-modal scientific data into a single training corpus, thereby potentially facilitating the development of foundation models for science. In this work, we introduce xVal, a strategy for continuously tokenizing numbers within language models that results in a more appropriate inductive bias for scientific applications. By training spec"},"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":"2310.02989","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-10-04T17:26:16Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"ba3dc882d4a0be07a88ce61b4f72acb90bb8aceefb5c01754a9ee92f22e59e31","abstract_canon_sha256":"d363a54128481748d5bc76566f31f6f97eaea939e42ee332b2b95dfeed6b75ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:49:00.417158Z","signature_b64":"RUT6/vMtt/Sl8newzY4VQCzGutzqVp0UIQ9y53Pe+IRVUQyG4ZbnE/bmYX8kXPp0pQGVFTv+SYvbW/y6mXEHBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"237ea3d81f5c37508805081711669a10bbefc2de8117adda8c9aa6de1b0a422e","last_reissued_at":"2026-07-05T09:49:00.416631Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:49:00.416631Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"xVal: A Continuous Numerical Tokenization for Scientific Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"stat.ML","authors_text":"Alberto Bietti, Bruno R\\'egaldo-Saint Blancard, Francois Lanusse, Geraud Krawezik, Kyunghyun Cho, Liam Parker, Mariel Pettee, Michael Eickenberg, Michael McCabe, Miles Cranmer, Ruben Ohana, Shirley Ho, Siavash Golkar, Tiberiu Tesileanu","submitted_at":"2023-10-04T17:26:16Z","abstract_excerpt":"Due in part to their discontinuous and discrete default encodings for numbers, Large Language Models (LLMs) have not yet been commonly used to process numerically-dense scientific datasets. Rendering datasets as text, however, could help aggregate diverse and multi-modal scientific data into a single training corpus, thereby potentially facilitating the development of foundation models for science. In this work, we introduce xVal, a strategy for continuously tokenizing numbers within language models that results in a more appropriate inductive bias for scientific applications. By training spec"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02989","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/2310.02989/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":"2310.02989","created_at":"2026-07-05T09:49:00.416689+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02989v2","created_at":"2026-07-05T09:49:00.416689+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02989","created_at":"2026-07-05T09:49:00.416689+00:00"},{"alias_kind":"pith_short_12","alias_value":"EN7KHWA7LQ3V","created_at":"2026-07-05T09:49:00.416689+00:00"},{"alias_kind":"pith_short_16","alias_value":"EN7KHWA7LQ3VBCAF","created_at":"2026-07-05T09:49:00.416689+00:00"},{"alias_kind":"pith_short_8","alias_value":"EN7KHWA7","created_at":"2026-07-05T09:49:00.416689+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18570","citing_title":"Streamlining Analysis and Design of Two-Dimensional Electronic Spectroscopy using Machine Learning","ref_index":157,"is_internal_anchor":false},{"citing_arxiv_id":"2502.09741","citing_title":"FoNE: Precise Single-Token Number Embeddings via Fourier Features","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2502.05564","citing_title":"TabICL: A Tabular Foundation Model for In-Context Learning on Large Data","ref_index":185,"is_internal_anchor":false},{"citing_arxiv_id":"2603.01444","citing_title":"Autoregressive Synthesis of Sparse and Semi-Structured Mixed-Type Data","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12106","citing_title":"Large Language Models as Amortized Pareto-Front Generators for Constrained Bi-Objective Convex Optimization","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11582","citing_title":"A Triadic Suffix Tokenization Scheme for Numerical Reasoning","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16775","citing_title":"Representation Before Training: A Fixed-Budget Benchmark for Generative Medical Event Models","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC","json":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC.json","graph_json":"https://pith.science/api/pith-number/EN7KHWA7LQ3VBCAFBALRCZU2CC/graph.json","events_json":"https://pith.science/api/pith-number/EN7KHWA7LQ3VBCAFBALRCZU2CC/events.json","paper":"https://pith.science/paper/EN7KHWA7"},"agent_actions":{"view_html":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC","download_json":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC.json","view_paper":"https://pith.science/paper/EN7KHWA7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02989&json=true","fetch_graph":"https://pith.science/api/pith-number/EN7KHWA7LQ3VBCAFBALRCZU2CC/graph.json","fetch_events":"https://pith.science/api/pith-number/EN7KHWA7LQ3VBCAFBALRCZU2CC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC/action/storage_attestation","attest_author":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC/action/author_attestation","sign_citation":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC/action/citation_signature","submit_replication":"https://pith.science/pith/EN7KHWA7LQ3VBCAFBALRCZU2CC/action/replication_record"}},"created_at":"2026-07-05T09:49:00.416689+00:00","updated_at":"2026-07-05T09:49:00.416689+00:00"}