{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CDFFE3Z6JVMBP3LLWGDBP5I4F5","short_pith_number":"pith:CDFFE3Z6","schema_version":"1.0","canonical_sha256":"10ca526f3e4d5817ed6bb18617f51c2f4e9b2fbfdebc0a53ce843f23fd66e38b","source":{"kind":"arxiv","id":"2605.19726","version":1},"attestation_state":"computed","paper":{"title":"Efficient Long-Context Modeling in Diffusion Language Models via Block Approximate Sparse Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hanbin Zhao, Huanyu Wang, Huanzhang Dou, Jiaya Jia, Senqiao Yang, Sitong Wu, Wenhu Zhang, YaoYang Liu, Yiming Wu","submitted_at":"2026-05-19T12:01:13Z","abstract_excerpt":"Diffusion Language Models (DLMs) enable globally coherent, bidirectional, and controllable text generation, offering advantages over traditional autoregressive LLMs, while scaling to ultra-long sequences remains costly. Many existing block-sparse attention methods select blocks by fixed sampling patterns over the high-resolution attention space, such as tail regions or anti-diagonal stripes. Such prior-driven sampling can miss salient tokens and introduce instability under distribution shifts. In this paper, we propose the Block Approximate Sparse Attention framework (BA-Att) with block-wise p"},"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":"2605.19726","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T12:01:13Z","cross_cats_sorted":[],"title_canon_sha256":"31ec88e960fc47728242d3a064b772e288f0326833e7132fea40e31b8a39f2a8","abstract_canon_sha256":"450318249d43b12e988cb99feaa1a0a39053a525e1b2faeca360b810886e848a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:06:11.180534Z","signature_b64":"DHoudIvtaYreo9HgNarvfALer+5lkDYCeBp2rAm8Kf1Kr4miZ6wWoD9gD+ukO/y89JUpiob+a8ozqxX6u5GLBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"10ca526f3e4d5817ed6bb18617f51c2f4e9b2fbfdebc0a53ce843f23fd66e38b","last_reissued_at":"2026-05-20T01:06:11.179991Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:06:11.179991Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Long-Context Modeling in Diffusion Language Models via Block Approximate Sparse Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hanbin Zhao, Huanyu Wang, Huanzhang Dou, Jiaya Jia, Senqiao Yang, Sitong Wu, Wenhu Zhang, YaoYang Liu, Yiming Wu","submitted_at":"2026-05-19T12:01:13Z","abstract_excerpt":"Diffusion Language Models (DLMs) enable globally coherent, bidirectional, and controllable text generation, offering advantages over traditional autoregressive LLMs, while scaling to ultra-long sequences remains costly. Many existing block-sparse attention methods select blocks by fixed sampling patterns over the high-resolution attention space, such as tail regions or anti-diagonal stripes. Such prior-driven sampling can miss salient tokens and introduce instability under distribution shifts. In this paper, we propose the Block Approximate Sparse Attention framework (BA-Att) with block-wise p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19726","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/2605.19726/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":"2605.19726","created_at":"2026-05-20T01:06:11.180059+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19726v1","created_at":"2026-05-20T01:06:11.180059+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19726","created_at":"2026-05-20T01:06:11.180059+00:00"},{"alias_kind":"pith_short_12","alias_value":"CDFFE3Z6JVMB","created_at":"2026-05-20T01:06:11.180059+00:00"},{"alias_kind":"pith_short_16","alias_value":"CDFFE3Z6JVMBP3LL","created_at":"2026-05-20T01:06:11.180059+00:00"},{"alias_kind":"pith_short_8","alias_value":"CDFFE3Z6","created_at":"2026-05-20T01:06:11.180059+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/CDFFE3Z6JVMBP3LLWGDBP5I4F5","json":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5.json","graph_json":"https://pith.science/api/pith-number/CDFFE3Z6JVMBP3LLWGDBP5I4F5/graph.json","events_json":"https://pith.science/api/pith-number/CDFFE3Z6JVMBP3LLWGDBP5I4F5/events.json","paper":"https://pith.science/paper/CDFFE3Z6"},"agent_actions":{"view_html":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5","download_json":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5.json","view_paper":"https://pith.science/paper/CDFFE3Z6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19726&json=true","fetch_graph":"https://pith.science/api/pith-number/CDFFE3Z6JVMBP3LLWGDBP5I4F5/graph.json","fetch_events":"https://pith.science/api/pith-number/CDFFE3Z6JVMBP3LLWGDBP5I4F5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5/action/storage_attestation","attest_author":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5/action/author_attestation","sign_citation":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5/action/citation_signature","submit_replication":"https://pith.science/pith/CDFFE3Z6JVMBP3LLWGDBP5I4F5/action/replication_record"}},"created_at":"2026-05-20T01:06:11.180059+00:00","updated_at":"2026-05-20T01:06:11.180059+00:00"}