{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:PZ45GQV4YPXTCRKN5HNIZ7N4EE","short_pith_number":"pith:PZ45GQV4","schema_version":"1.0","canonical_sha256":"7e79d342bcc3ef31454de9da8cfdbc211359100fa5509ec82b8b5c41e74367a5","source":{"kind":"arxiv","id":"2309.06236","version":1},"attestation_state":"computed","paper":{"title":"The first step is the hardest: Pitfalls of Representing and Tokenizing Temporal Data for Large Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dimitris Spathis, Fahim Kawsar","submitted_at":"2023-09-12T13:51:29Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable generalization across diverse tasks, leading individuals to increasingly use them as personal assistants and universal computing engines. Nevertheless, a notable obstacle emerges when feeding numerical/temporal data into these models, such as data sourced from wearables or electronic health records. LLMs employ tokenizers in their input that break down text into smaller units. However, tokenizers are not designed to represent numerical values and might struggle to understand repetitive patterns and context, treating consecutive values a"},"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":"2309.06236","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2023-09-12T13:51:29Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"187f68af0b013d150ba8042665d4f541bbd4b4eb2fc856a7691e2cd17d66228b","abstract_canon_sha256":"ab2021fc9fb1c5c1242d729cdd4eaa1f275e33eff9c72e82754f7341c41fa55f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:50:05.328703Z","signature_b64":"6XNXh3H5WmX6xL3SNa8vbcNU3ddt0pqeQzaeFhOXdG+d2kV9Zqp6HP/olmBV8yPjxdSp/OoTkj7vwHYf84XZDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7e79d342bcc3ef31454de9da8cfdbc211359100fa5509ec82b8b5c41e74367a5","last_reissued_at":"2026-07-05T06:50:05.328206Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:50:05.328206Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The first step is the hardest: Pitfalls of Representing and Tokenizing Temporal Data for Large Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dimitris Spathis, Fahim Kawsar","submitted_at":"2023-09-12T13:51:29Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable generalization across diverse tasks, leading individuals to increasingly use them as personal assistants and universal computing engines. Nevertheless, a notable obstacle emerges when feeding numerical/temporal data into these models, such as data sourced from wearables or electronic health records. LLMs employ tokenizers in their input that break down text into smaller units. However, tokenizers are not designed to represent numerical values and might struggle to understand repetitive patterns and context, treating consecutive values a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.06236","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/2309.06236/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":"2309.06236","created_at":"2026-07-05T06:50:05.328265+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.06236v1","created_at":"2026-07-05T06:50:05.328265+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.06236","created_at":"2026-07-05T06:50:05.328265+00:00"},{"alias_kind":"pith_short_12","alias_value":"PZ45GQV4YPXT","created_at":"2026-07-05T06:50:05.328265+00:00"},{"alias_kind":"pith_short_16","alias_value":"PZ45GQV4YPXTCRKN","created_at":"2026-07-05T06:50:05.328265+00:00"},{"alias_kind":"pith_short_8","alias_value":"PZ45GQV4","created_at":"2026-07-05T06:50:05.328265+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2511.01101","citing_title":"TSVer: A Benchmark for Fact Verification Against Time-Series Evidence","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02711","citing_title":"Foundation Models Defining A New Era In Sensor-based Human Activity Recognition: A Survey And Outlook","ref_index":134,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE","json":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE.json","graph_json":"https://pith.science/api/pith-number/PZ45GQV4YPXTCRKN5HNIZ7N4EE/graph.json","events_json":"https://pith.science/api/pith-number/PZ45GQV4YPXTCRKN5HNIZ7N4EE/events.json","paper":"https://pith.science/paper/PZ45GQV4"},"agent_actions":{"view_html":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE","download_json":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE.json","view_paper":"https://pith.science/paper/PZ45GQV4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.06236&json=true","fetch_graph":"https://pith.science/api/pith-number/PZ45GQV4YPXTCRKN5HNIZ7N4EE/graph.json","fetch_events":"https://pith.science/api/pith-number/PZ45GQV4YPXTCRKN5HNIZ7N4EE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE/action/storage_attestation","attest_author":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE/action/author_attestation","sign_citation":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE/action/citation_signature","submit_replication":"https://pith.science/pith/PZ45GQV4YPXTCRKN5HNIZ7N4EE/action/replication_record"}},"created_at":"2026-07-05T06:50:05.328265+00:00","updated_at":"2026-07-05T06:50:05.328265+00:00"}