{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:HVNTKDJRWHZMVTCSZH2GX6QKQZ","short_pith_number":"pith:HVNTKDJR","schema_version":"1.0","canonical_sha256":"3d5b350d31b1f2cacc52c9f46bfa0a86761b743692ab25a875cf1f51d28ef655","source":{"kind":"arxiv","id":"2305.16340","version":3},"attestation_state":"computed","paper":{"title":"Segmented Recurrent Transformer: An Efficient Sequence-to-Sequence Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kaushik Roy, Sayeed Shafayet Chowdhury, Yinghan Long","submitted_at":"2023-05-24T03:47:22Z","abstract_excerpt":"Transformers have shown dominant performance across a range of domains including language and vision. However, their computational cost grows quadratically with the sequence length, making their usage prohibitive for resource-constrained applications. To counter this, our approach is to divide the whole sequence into segments and apply attention to the individual segments. We propose a segmented recurrent transformer (SRformer) that combines segmented (local) attention with recurrent attention. The loss caused by reducing the attention window length is compensated by aggregating information ac"},"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":"2305.16340","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T03:47:22Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"f08d7aa1938f5777d591f00aa2746e4be8d0680e26d0cd780817fb5245df3385","abstract_canon_sha256":"348e7d91392ff6e020c6629aebed78ebe783963dc5f8693f5a4abfccb28d4ff6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:03:26.133436Z","signature_b64":"zNyS2rS/ERhLBzlNOaVeSv8TdaNrx7V3YxYtwHzNEx0cie6XcSxW5+Hpyci6MpJ4CsQnu0WrEm6zlqRrrPr6Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3d5b350d31b1f2cacc52c9f46bfa0a86761b743692ab25a875cf1f51d28ef655","last_reissued_at":"2026-07-05T07:03:26.132971Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:03:26.132971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Segmented Recurrent Transformer: An Efficient Sequence-to-Sequence Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kaushik Roy, Sayeed Shafayet Chowdhury, Yinghan Long","submitted_at":"2023-05-24T03:47:22Z","abstract_excerpt":"Transformers have shown dominant performance across a range of domains including language and vision. However, their computational cost grows quadratically with the sequence length, making their usage prohibitive for resource-constrained applications. To counter this, our approach is to divide the whole sequence into segments and apply attention to the individual segments. We propose a segmented recurrent transformer (SRformer) that combines segmented (local) attention with recurrent attention. The loss caused by reducing the attention window length is compensated by aggregating information ac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.16340","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/2305.16340/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":"2305.16340","created_at":"2026-07-05T07:03:26.133023+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.16340v3","created_at":"2026-07-05T07:03:26.133023+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.16340","created_at":"2026-07-05T07:03:26.133023+00:00"},{"alias_kind":"pith_short_12","alias_value":"HVNTKDJRWHZM","created_at":"2026-07-05T07:03:26.133023+00:00"},{"alias_kind":"pith_short_16","alias_value":"HVNTKDJRWHZMVTCS","created_at":"2026-07-05T07:03:26.133023+00:00"},{"alias_kind":"pith_short_8","alias_value":"HVNTKDJR","created_at":"2026-07-05T07:03:26.133023+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.06139","citing_title":"LCIRC: A Recurrent Compression Approach for Efficient Long-form Context and Query Dependent Modeling in LLMs","ref_index":28,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ","json":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ.json","graph_json":"https://pith.science/api/pith-number/HVNTKDJRWHZMVTCSZH2GX6QKQZ/graph.json","events_json":"https://pith.science/api/pith-number/HVNTKDJRWHZMVTCSZH2GX6QKQZ/events.json","paper":"https://pith.science/paper/HVNTKDJR"},"agent_actions":{"view_html":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ","download_json":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ.json","view_paper":"https://pith.science/paper/HVNTKDJR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.16340&json=true","fetch_graph":"https://pith.science/api/pith-number/HVNTKDJRWHZMVTCSZH2GX6QKQZ/graph.json","fetch_events":"https://pith.science/api/pith-number/HVNTKDJRWHZMVTCSZH2GX6QKQZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ/action/storage_attestation","attest_author":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ/action/author_attestation","sign_citation":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ/action/citation_signature","submit_replication":"https://pith.science/pith/HVNTKDJRWHZMVTCSZH2GX6QKQZ/action/replication_record"}},"created_at":"2026-07-05T07:03:26.133023+00:00","updated_at":"2026-07-05T07:03:26.133023+00:00"}