{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BI3CC33TPYZRPAKFFQZQK2B6MA","short_pith_number":"pith:BI3CC33T","schema_version":"1.0","canonical_sha256":"0a36216f737e331781452c3305683e60228fbc19bf3b1612599efbec845ea1ac","source":{"kind":"arxiv","id":"2601.09566","version":4},"attestation_state":"computed","paper":{"title":"Hot-Start Chinese Language Modeling:Visual Glyphs Accelerate Sample-Efficient Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hao Guan, Shuyang Xiang","submitted_at":"2026-01-14T15:34:37Z","abstract_excerpt":"In this work, we study whether rendering Chinese characters as visual glyph images, rather than discrete token IDs as mainstream LLMs do, providing an inductive bias for character-level language modeling. Our central finding gives a double-edged insight: visual inputs produce a pronounced hot-start effect, more than doubling early-stage accuracy within the first epoch (at 0.4% of total training steps) (12.3% visual inputs vs. 5.8% index-based baseline), yet both approaches converge to essentially identical final accuracy (39%). This pattern holds across resolutions as low as 8x8 pixels, partia"},"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":"2601.09566","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-14T15:34:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5ed35e2da6d39d76d3902029e7a5fb2de9167608619bd42c6bcf222c0c4450ab","abstract_canon_sha256":"652bd53041fa3a1ced27467947f69750051d00c212b127f63a7f16a30d20f7a1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:41.003191Z","signature_b64":"stTrMg3Cn6wG9AzklrUr2tijElUANjV45nJczYAOsWJg7zHzBICCAQxUOCJUWBIdtMUDfduc2GPRob0wVAr/DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0a36216f737e331781452c3305683e60228fbc19bf3b1612599efbec845ea1ac","last_reissued_at":"2026-06-02T01:03:41.002627Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:41.002627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Hot-Start Chinese Language Modeling:Visual Glyphs Accelerate Sample-Efficient Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hao Guan, Shuyang Xiang","submitted_at":"2026-01-14T15:34:37Z","abstract_excerpt":"In this work, we study whether rendering Chinese characters as visual glyph images, rather than discrete token IDs as mainstream LLMs do, providing an inductive bias for character-level language modeling. Our central finding gives a double-edged insight: visual inputs produce a pronounced hot-start effect, more than doubling early-stage accuracy within the first epoch (at 0.4% of total training steps) (12.3% visual inputs vs. 5.8% index-based baseline), yet both approaches converge to essentially identical final accuracy (39%). This pattern holds across resolutions as low as 8x8 pixels, partia"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.09566","kind":"arxiv","version":4},"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/2601.09566/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":"2601.09566","created_at":"2026-06-02T01:03:41.002699+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.09566v4","created_at":"2026-06-02T01:03:41.002699+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.09566","created_at":"2026-06-02T01:03:41.002699+00:00"},{"alias_kind":"pith_short_12","alias_value":"BI3CC33TPYZR","created_at":"2026-06-02T01:03:41.002699+00:00"},{"alias_kind":"pith_short_16","alias_value":"BI3CC33TPYZRPAKF","created_at":"2026-06-02T01:03:41.002699+00:00"},{"alias_kind":"pith_short_8","alias_value":"BI3CC33T","created_at":"2026-06-02T01:03:41.002699+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA","json":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA.json","graph_json":"https://pith.science/api/pith-number/BI3CC33TPYZRPAKFFQZQK2B6MA/graph.json","events_json":"https://pith.science/api/pith-number/BI3CC33TPYZRPAKFFQZQK2B6MA/events.json","paper":"https://pith.science/paper/BI3CC33T"},"agent_actions":{"view_html":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA","download_json":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA.json","view_paper":"https://pith.science/paper/BI3CC33T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.09566&json=true","fetch_graph":"https://pith.science/api/pith-number/BI3CC33TPYZRPAKFFQZQK2B6MA/graph.json","fetch_events":"https://pith.science/api/pith-number/BI3CC33TPYZRPAKFFQZQK2B6MA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA/action/storage_attestation","attest_author":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA/action/author_attestation","sign_citation":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA/action/citation_signature","submit_replication":"https://pith.science/pith/BI3CC33TPYZRPAKFFQZQK2B6MA/action/replication_record"}},"created_at":"2026-06-02T01:03:41.002699+00:00","updated_at":"2026-06-02T01:03:41.002699+00:00"}