{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:VFO27VEDWT2X7T7CEOWMM2NKEC","short_pith_number":"pith:VFO27VED","schema_version":"1.0","canonical_sha256":"a95dafd483b4f57fcfe223acc669aa208e818fd5a9c7d54d1d29fba8344cb07b","source":{"kind":"arxiv","id":"2105.06679","version":2},"attestation_state":"computed","paper":{"title":"Dynamic Multi-Branch Layers for On-Device Neural Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Maosong Sun, Meng Zhang, Qun Liu, Yang Liu, Zeyuan Yang, Zhixing Tan","submitted_at":"2021-05-14T07:32:53Z","abstract_excerpt":"With the rapid development of artificial intelligence (AI), there is a trend in moving AI applications, such as neural machine translation (NMT), from cloud to mobile devices. Constrained by limited hardware resources and battery, the performance of on-device NMT systems is far from satisfactory. Inspired by conditional computation, we propose to improve the performance of on-device NMT systems with dynamic multi-branch layers. Specifically, we design a layer-wise dynamic multi-branch network with only one branch activated during training and inference. As not all branches are activated during"},"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":"2105.06679","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-05-14T07:32:53Z","cross_cats_sorted":[],"title_canon_sha256":"42693619a250aea08ef96e42dd42547aae428204222d28090d082cb42854c595","abstract_canon_sha256":"07470941c9bd499d260332b7fa46df5ae5a125747addf4425e38abe743e1075d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:05:58.460235Z","signature_b64":"o+jdiEWtcYLmicCKj1fQyuy3EqXFq2frw7oeGsYV69I24YxuE9iP/RoY+tLRTva2hskTWkD9VtqrOg7aOKudDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a95dafd483b4f57fcfe223acc669aa208e818fd5a9c7d54d1d29fba8344cb07b","last_reissued_at":"2026-07-05T04:05:58.459824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:05:58.459824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamic Multi-Branch Layers for On-Device Neural Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Maosong Sun, Meng Zhang, Qun Liu, Yang Liu, Zeyuan Yang, Zhixing Tan","submitted_at":"2021-05-14T07:32:53Z","abstract_excerpt":"With the rapid development of artificial intelligence (AI), there is a trend in moving AI applications, such as neural machine translation (NMT), from cloud to mobile devices. Constrained by limited hardware resources and battery, the performance of on-device NMT systems is far from satisfactory. Inspired by conditional computation, we propose to improve the performance of on-device NMT systems with dynamic multi-branch layers. Specifically, we design a layer-wise dynamic multi-branch network with only one branch activated during training and inference. As not all branches are activated during"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.06679","kind":"arxiv","version":2},"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/2105.06679/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":"2105.06679","created_at":"2026-07-05T04:05:58.459879+00:00"},{"alias_kind":"arxiv_version","alias_value":"2105.06679v2","created_at":"2026-07-05T04:05:58.459879+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.06679","created_at":"2026-07-05T04:05:58.459879+00:00"},{"alias_kind":"pith_short_12","alias_value":"VFO27VEDWT2X","created_at":"2026-07-05T04:05:58.459879+00:00"},{"alias_kind":"pith_short_16","alias_value":"VFO27VEDWT2X7T7C","created_at":"2026-07-05T04:05:58.459879+00:00"},{"alias_kind":"pith_short_8","alias_value":"VFO27VED","created_at":"2026-07-05T04:05:58.459879+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/VFO27VEDWT2X7T7CEOWMM2NKEC","json":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC.json","graph_json":"https://pith.science/api/pith-number/VFO27VEDWT2X7T7CEOWMM2NKEC/graph.json","events_json":"https://pith.science/api/pith-number/VFO27VEDWT2X7T7CEOWMM2NKEC/events.json","paper":"https://pith.science/paper/VFO27VED"},"agent_actions":{"view_html":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC","download_json":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC.json","view_paper":"https://pith.science/paper/VFO27VED","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2105.06679&json=true","fetch_graph":"https://pith.science/api/pith-number/VFO27VEDWT2X7T7CEOWMM2NKEC/graph.json","fetch_events":"https://pith.science/api/pith-number/VFO27VEDWT2X7T7CEOWMM2NKEC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC/action/storage_attestation","attest_author":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC/action/author_attestation","sign_citation":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC/action/citation_signature","submit_replication":"https://pith.science/pith/VFO27VEDWT2X7T7CEOWMM2NKEC/action/replication_record"}},"created_at":"2026-07-05T04:05:58.459879+00:00","updated_at":"2026-07-05T04:05:58.459879+00:00"}