{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6HIISJWNFRNI3VTGNLF35ET7IK","short_pith_number":"pith:6HIISJWN","schema_version":"1.0","canonical_sha256":"f1d08926cd2c5a8dd6666acbbe927f429659870dedefa854ed2b4d9f84e0a7df","source":{"kind":"arxiv","id":"2603.16077","version":3},"attestation_state":"computed","paper":{"title":"MDM-Prime-v2: Binary Encoding and Index Shuffling Enable Scaling of Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chen-Hao Chao, Chun-Yi Lee, Junwei Quan, Rahul G. Krishnan, Wei-Fang Sun","submitted_at":"2026-03-17T02:54:16Z","abstract_excerpt":"Masked diffusion models (MDM) exhibit superior generalization when learned using a Partial masking scheme (Prime). This approach converts tokens into sub-tokens and models the diffusion process at the sub-token level. We identify two limitations of the MDM-Prime framework. First, we find that the functional form of the subtokenizer significantly increases the cross-entropy loss in the objective when paired with commonly used Byte-Pair-Encoding (BPE) tokenizers. Second, we lack tools to guide the hyperparameter choice of the token granularity in the subtokenizer. To address these limitations, w"},"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":"2603.16077","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T02:54:16Z","cross_cats_sorted":[],"title_canon_sha256":"232dc460b5a450d5ab24e1f34f88dc25201ba622f5956300701ca1b93a866680","abstract_canon_sha256":"8abea9dc14e850646ce813eeaa6f806ed4c58f4493e8fe3d9fa7451437f961ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:00.370462Z","signature_b64":"4zBABBoPn/IkV6I2RVJjMtio8n2iOgSg5vg0dN/u7b1vnYWU8bPSFslaHseo4wqXjXHQtZSdgePNNxuFJL6eAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f1d08926cd2c5a8dd6666acbbe927f429659870dedefa854ed2b4d9f84e0a7df","last_reissued_at":"2026-05-22T01:04:00.369794Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:00.369794Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MDM-Prime-v2: Binary Encoding and Index Shuffling Enable Scaling of Diffusion Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chen-Hao Chao, Chun-Yi Lee, Junwei Quan, Rahul G. Krishnan, Wei-Fang Sun","submitted_at":"2026-03-17T02:54:16Z","abstract_excerpt":"Masked diffusion models (MDM) exhibit superior generalization when learned using a Partial masking scheme (Prime). This approach converts tokens into sub-tokens and models the diffusion process at the sub-token level. We identify two limitations of the MDM-Prime framework. First, we find that the functional form of the subtokenizer significantly increases the cross-entropy loss in the objective when paired with commonly used Byte-Pair-Encoding (BPE) tokenizers. Second, we lack tools to guide the hyperparameter choice of the token granularity in the subtokenizer. To address these limitations, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.16077","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/2603.16077/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":"2603.16077","created_at":"2026-05-22T01:04:00.369883+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.16077v3","created_at":"2026-05-22T01:04:00.369883+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.16077","created_at":"2026-05-22T01:04:00.369883+00:00"},{"alias_kind":"pith_short_12","alias_value":"6HIISJWNFRNI","created_at":"2026-05-22T01:04:00.369883+00:00"},{"alias_kind":"pith_short_16","alias_value":"6HIISJWNFRNI3VTG","created_at":"2026-05-22T01:04:00.369883+00:00"},{"alias_kind":"pith_short_8","alias_value":"6HIISJWN","created_at":"2026-05-22T01:04:00.369883+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/6HIISJWNFRNI3VTGNLF35ET7IK","json":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK.json","graph_json":"https://pith.science/api/pith-number/6HIISJWNFRNI3VTGNLF35ET7IK/graph.json","events_json":"https://pith.science/api/pith-number/6HIISJWNFRNI3VTGNLF35ET7IK/events.json","paper":"https://pith.science/paper/6HIISJWN"},"agent_actions":{"view_html":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK","download_json":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK.json","view_paper":"https://pith.science/paper/6HIISJWN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.16077&json=true","fetch_graph":"https://pith.science/api/pith-number/6HIISJWNFRNI3VTGNLF35ET7IK/graph.json","fetch_events":"https://pith.science/api/pith-number/6HIISJWNFRNI3VTGNLF35ET7IK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK/action/storage_attestation","attest_author":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK/action/author_attestation","sign_citation":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK/action/citation_signature","submit_replication":"https://pith.science/pith/6HIISJWNFRNI3VTGNLF35ET7IK/action/replication_record"}},"created_at":"2026-05-22T01:04:00.369883+00:00","updated_at":"2026-05-22T01:04:00.369883+00:00"}