{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:UX4SGRUFQXVK32SVDKF4ESWMY5","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"0d9ae6d7882f8b9b0360977a3904a207bed42fb5a9d85e9c9098b310a7b4bf16","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-07T16:25:27Z","title_canon_sha256":"c43f1dff7516e59a2951f2384cb296312a7e91c33a53abbf9e8d704b6bed4fc4"},"schema_version":"1.0","source":{"id":"2106.03760","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.03760","created_at":"2026-07-05T03:44:54Z"},{"alias_kind":"arxiv_version","alias_value":"2106.03760v3","created_at":"2026-07-05T03:44:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.03760","created_at":"2026-07-05T03:44:54Z"},{"alias_kind":"pith_short_12","alias_value":"UX4SGRUFQXVK","created_at":"2026-07-05T03:44:54Z"},{"alias_kind":"pith_short_16","alias_value":"UX4SGRUFQXVK32SV","created_at":"2026-07-05T03:44:54Z"},{"alias_kind":"pith_short_8","alias_value":"UX4SGRUF","created_at":"2026-07-05T03:44:54Z"}],"graph_snapshots":[{"event_id":"sha256:4943e61df2f7d582854358a11bf0008651a47105daddbbabbee5f8d49f3b0161","target":"graph","created_at":"2026-07-05T03:44:54Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2106.03760/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The Mixture-of-Experts (MoE) architecture is showing promising results in improving parameter sharing in multi-task learning (MTL) and in scaling high-capacity neural networks. State-of-the-art MoE models use a trainable sparse gate to select a subset of the experts for each input example. While conceptually appealing, existing sparse gates, such as Top-k, are not smooth. The lack of smoothness can lead to convergence and statistical performance issues when training with gradient-based methods. In this paper, we develop DSelect-k: a continuously differentiable and sparse gate for MoE, based on","authors_text":"Aakanksha Chowdhery, Ed H. Chi, Hussein Hazimeh, Lichan Hong, Maheswaran Sathiamoorthy, Rahul Mazumder, Yihua Chen, Zhe Zhao","cross_cats":["math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-07T16:25:27Z","title":"DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.03760","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:becf8b4572e90ee8c0833c7c480d90f254dd537b447654e2acb7c85b0d9edd15","target":"record","created_at":"2026-07-05T03:44:54Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"0d9ae6d7882f8b9b0360977a3904a207bed42fb5a9d85e9c9098b310a7b4bf16","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-07T16:25:27Z","title_canon_sha256":"c43f1dff7516e59a2951f2384cb296312a7e91c33a53abbf9e8d704b6bed4fc4"},"schema_version":"1.0","source":{"id":"2106.03760","kind":"arxiv","version":3}},"canonical_sha256":"a5f923468585eaadea551a8bc24accc75e0370420d8e414f3b82b967c6af01c5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a5f923468585eaadea551a8bc24accc75e0370420d8e414f3b82b967c6af01c5","first_computed_at":"2026-07-05T03:44:54.033040Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:44:54.033040Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Z6XUNNTJ4xaLfOdQnTfqUDBQx/ymLYnenWBlggOUBB47aGOokWbkqoGCP9oOguAeVG0a17qtjQQlESpivs0vBg==","signature_status":"signed_v1","signed_at":"2026-07-05T03:44:54.033495Z","signed_message":"canonical_sha256_bytes"},"source_id":"2106.03760","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:becf8b4572e90ee8c0833c7c480d90f254dd537b447654e2acb7c85b0d9edd15","sha256:4943e61df2f7d582854358a11bf0008651a47105daddbbabbee5f8d49f3b0161"],"state_sha256":"907c2448620bacb965d5826a860ed0fc4ab73af89041c1b35ffc17acb5808c55"}