{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:42Q3QQTWC2O2JL6HGCA2DVJGJG","short_pith_number":"pith:42Q3QQTW","schema_version":"1.0","canonical_sha256":"e6a1b84276169da4afc73081a1d526499931b9c4d0238376e0d95c7307e95569","source":{"kind":"arxiv","id":"2006.08063","version":3},"attestation_state":"computed","paper":{"title":"Gradient Estimation with Stochastic Softmax Tricks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Andreas Krause, Chris J. Maddison, Dami Choi, Daniel Tarlow, Max B. Paulus","submitted_at":"2020-06-15T00:43:44Z","abstract_excerpt":"The Gumbel-Max trick is the basis of many relaxed gradient estimators. These estimators are easy to implement and low variance, but the goal of scaling them comprehensively to large combinatorial distributions is still outstanding. Working within the perturbation model framework, we introduce stochastic softmax tricks, which generalize the Gumbel-Softmax trick to combinatorial spaces. Our framework is a unified perspective on existing relaxed estimators for perturbation models, and it contains many novel relaxations. We design structured relaxations for subset selection, spanning trees, arbore"},"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":"2006.08063","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-06-15T00:43:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7572f2b1888e50ecbfdd16a9cc0f9b8a410c1388637046a5dc6e8c92f269e506","abstract_canon_sha256":"3e42201542214349f448d3850c116a7f03d0bdadf203c8308ba667b598190e23"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:18:52.551630Z","signature_b64":"lfbo9SVOKRketg9azIucJjr1Mz53lud4PKg1knljFNNvqP851yHhjDFegeR/1ag0rL/6hd875dkUQQPZbkdMBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e6a1b84276169da4afc73081a1d526499931b9c4d0238376e0d95c7307e95569","last_reissued_at":"2026-07-05T02:18:52.551144Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:18:52.551144Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Estimation with Stochastic Softmax Tricks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Andreas Krause, Chris J. Maddison, Dami Choi, Daniel Tarlow, Max B. Paulus","submitted_at":"2020-06-15T00:43:44Z","abstract_excerpt":"The Gumbel-Max trick is the basis of many relaxed gradient estimators. These estimators are easy to implement and low variance, but the goal of scaling them comprehensively to large combinatorial distributions is still outstanding. Working within the perturbation model framework, we introduce stochastic softmax tricks, which generalize the Gumbel-Softmax trick to combinatorial spaces. Our framework is a unified perspective on existing relaxed estimators for perturbation models, and it contains many novel relaxations. We design structured relaxations for subset selection, spanning trees, arbore"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.08063","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/2006.08063/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":"2006.08063","created_at":"2026-07-05T02:18:52.551203+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.08063v3","created_at":"2026-07-05T02:18:52.551203+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.08063","created_at":"2026-07-05T02:18:52.551203+00:00"},{"alias_kind":"pith_short_12","alias_value":"42Q3QQTWC2O2","created_at":"2026-07-05T02:18:52.551203+00:00"},{"alias_kind":"pith_short_16","alias_value":"42Q3QQTWC2O2JL6H","created_at":"2026-07-05T02:18:52.551203+00:00"},{"alias_kind":"pith_short_8","alias_value":"42Q3QQTW","created_at":"2026-07-05T02:18:52.551203+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/42Q3QQTWC2O2JL6HGCA2DVJGJG","json":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG.json","graph_json":"https://pith.science/api/pith-number/42Q3QQTWC2O2JL6HGCA2DVJGJG/graph.json","events_json":"https://pith.science/api/pith-number/42Q3QQTWC2O2JL6HGCA2DVJGJG/events.json","paper":"https://pith.science/paper/42Q3QQTW"},"agent_actions":{"view_html":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG","download_json":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG.json","view_paper":"https://pith.science/paper/42Q3QQTW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.08063&json=true","fetch_graph":"https://pith.science/api/pith-number/42Q3QQTWC2O2JL6HGCA2DVJGJG/graph.json","fetch_events":"https://pith.science/api/pith-number/42Q3QQTWC2O2JL6HGCA2DVJGJG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG/action/storage_attestation","attest_author":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG/action/author_attestation","sign_citation":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG/action/citation_signature","submit_replication":"https://pith.science/pith/42Q3QQTWC2O2JL6HGCA2DVJGJG/action/replication_record"}},"created_at":"2026-07-05T02:18:52.551203+00:00","updated_at":"2026-07-05T02:18:52.551203+00:00"}