{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2017:6B23RVQ6VHK2DGWDJS6X3W7BOT","short_pith_number":"pith:6B23RVQ6","canonical_record":{"source":{"id":"1710.09026","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-10-25T00:20:55Z","cross_cats_sorted":["cs.CL","eess.AS","stat.ML"],"title_canon_sha256":"98929081537876fdef071fa82e1821854bf13f239bb5b5ee5e5566c79bda8cab","abstract_canon_sha256":"0b4b05293aee9b210e5aa5a56a2ed588c00fe3fdff122a9f9fca8d1eb69d22ae"},"schema_version":"1.0"},"canonical_sha256":"f075b8d61ea9d5a19ac34cbd7ddbe174fe569df84d493e23f35deeaf024251cc","source":{"kind":"arxiv","id":"1710.09026","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1710.09026","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"arxiv_version","alias_value":"1710.09026v2","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1710.09026","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"pith_short_12","alias_value":"6B23RVQ6VHK2","created_at":"2026-05-18T12:31:03Z"},{"alias_kind":"pith_short_16","alias_value":"6B23RVQ6VHK2DGWD","created_at":"2026-05-18T12:31:03Z"},{"alias_kind":"pith_short_8","alias_value":"6B23RVQ6","created_at":"2026-05-18T12:31:03Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2017:6B23RVQ6VHK2DGWDJS6X3W7BOT","target":"record","payload":{"canonical_record":{"source":{"id":"1710.09026","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-10-25T00:20:55Z","cross_cats_sorted":["cs.CL","eess.AS","stat.ML"],"title_canon_sha256":"98929081537876fdef071fa82e1821854bf13f239bb5b5ee5e5566c79bda8cab","abstract_canon_sha256":"0b4b05293aee9b210e5aa5a56a2ed588c00fe3fdff122a9f9fca8d1eb69d22ae"},"schema_version":"1.0"},"canonical_sha256":"f075b8d61ea9d5a19ac34cbd7ddbe174fe569df84d493e23f35deeaf024251cc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:24:22.276960Z","signature_b64":"cOImGGcqLXC17eyOcFzJpwKvncE1g5LaAKqqYtsecGI3LchIGxZlV99LwKHM/2/W3feAV4JlKppfkBZX+ONfDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f075b8d61ea9d5a19ac34cbd7ddbe174fe569df84d493e23f35deeaf024251cc","last_reissued_at":"2026-05-18T00:24:22.276222Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:24:22.276222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1710.09026","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:24:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"L+wAWZ3RN7GQLey820Lph92ZfGyBhp/fvn/fVn1Et6rc5qTqJNhJcxth+HD9559akGjvW2SwT1+T/r0DLOD2BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T15:13:32.246331Z"},"content_sha256":"540da2ac0f02e29b46a54868ce568dae0e7e7bc725f3a54e4dcebd77ea7f64c9","schema_version":"1.0","event_id":"sha256:540da2ac0f02e29b46a54868ce568dae0e7e7bc725f3a54e4dcebd77ea7f64c9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2017:6B23RVQ6VHK2DGWDJS6X3W7BOT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Trace norm regularization and faster inference for embedded speech recognition RNNs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","eess.AS","stat.ML"],"primary_cat":"cs.LG","authors_text":"Kavya Srinet, Kexin Zhao, Markus Kliegl, Mohammad Shoeybi, Siddharth Goyal","submitted_at":"2017-10-25T00:20:55Z","abstract_excerpt":"We propose and evaluate new techniques for compressing and speeding up dense matrix multiplications as found in the fully connected and recurrent layers of neural networks for embedded large vocabulary continuous speech recognition (LVCSR). For compression, we introduce and study a trace norm regularization technique for training low rank factored versions of matrix multiplications. Compared to standard low rank training, we show that our method leads to good accuracy versus number of parameter trade-offs and can be used to speed up training of large models. For speedup, we enable faster infer"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1710.09026","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":""},"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:24:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pE5viU12+MZREGYlKrAHUUgnuRuvhJHajOA2j5M58ydb4Od7gctMtQxa2GxKOjW32mb1VGm54kzOQTsnUjr3Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T15:13:32.246732Z"},"content_sha256":"73130839f4817edd99d49c90254718516dd5d3879d174de7f358ce46cf315a33","schema_version":"1.0","event_id":"sha256:73130839f4817edd99d49c90254718516dd5d3879d174de7f358ce46cf315a33"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/bundle.json","state_url":"https://pith.science/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-29T15:13:32Z","links":{"resolver":"https://pith.science/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT","bundle":"https://pith.science/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/bundle.json","state":"https://pith.science/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6B23RVQ6VHK2DGWDJS6X3W7BOT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2017:6B23RVQ6VHK2DGWDJS6X3W7BOT","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":"0b4b05293aee9b210e5aa5a56a2ed588c00fe3fdff122a9f9fca8d1eb69d22ae","cross_cats_sorted":["cs.CL","eess.AS","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-10-25T00:20:55Z","title_canon_sha256":"98929081537876fdef071fa82e1821854bf13f239bb5b5ee5e5566c79bda8cab"},"schema_version":"1.0","source":{"id":"1710.09026","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1710.09026","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"arxiv_version","alias_value":"1710.09026v2","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1710.09026","created_at":"2026-05-18T00:24:22Z"},{"alias_kind":"pith_short_12","alias_value":"6B23RVQ6VHK2","created_at":"2026-05-18T12:31:03Z"},{"alias_kind":"pith_short_16","alias_value":"6B23RVQ6VHK2DGWD","created_at":"2026-05-18T12:31:03Z"},{"alias_kind":"pith_short_8","alias_value":"6B23RVQ6","created_at":"2026-05-18T12:31:03Z"}],"graph_snapshots":[{"event_id":"sha256:73130839f4817edd99d49c90254718516dd5d3879d174de7f358ce46cf315a33","target":"graph","created_at":"2026-05-18T00:24:22Z","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"},"paper":{"abstract_excerpt":"We propose and evaluate new techniques for compressing and speeding up dense matrix multiplications as found in the fully connected and recurrent layers of neural networks for embedded large vocabulary continuous speech recognition (LVCSR). For compression, we introduce and study a trace norm regularization technique for training low rank factored versions of matrix multiplications. Compared to standard low rank training, we show that our method leads to good accuracy versus number of parameter trade-offs and can be used to speed up training of large models. For speedup, we enable faster infer","authors_text":"Kavya Srinet, Kexin Zhao, Markus Kliegl, Mohammad Shoeybi, Siddharth Goyal","cross_cats":["cs.CL","eess.AS","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-10-25T00:20:55Z","title":"Trace norm regularization and faster inference for embedded speech recognition RNNs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1710.09026","kind":"arxiv","version":2},"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":"sha256:540da2ac0f02e29b46a54868ce568dae0e7e7bc725f3a54e4dcebd77ea7f64c9","target":"record","created_at":"2026-05-18T00:24:22Z","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":"0b4b05293aee9b210e5aa5a56a2ed588c00fe3fdff122a9f9fca8d1eb69d22ae","cross_cats_sorted":["cs.CL","eess.AS","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-10-25T00:20:55Z","title_canon_sha256":"98929081537876fdef071fa82e1821854bf13f239bb5b5ee5e5566c79bda8cab"},"schema_version":"1.0","source":{"id":"1710.09026","kind":"arxiv","version":2}},"canonical_sha256":"f075b8d61ea9d5a19ac34cbd7ddbe174fe569df84d493e23f35deeaf024251cc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f075b8d61ea9d5a19ac34cbd7ddbe174fe569df84d493e23f35deeaf024251cc","first_computed_at":"2026-05-18T00:24:22.276222Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:24:22.276222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cOImGGcqLXC17eyOcFzJpwKvncE1g5LaAKqqYtsecGI3LchIGxZlV99LwKHM/2/W3feAV4JlKppfkBZX+ONfDA==","signature_status":"signed_v1","signed_at":"2026-05-18T00:24:22.276960Z","signed_message":"canonical_sha256_bytes"},"source_id":"1710.09026","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:540da2ac0f02e29b46a54868ce568dae0e7e7bc725f3a54e4dcebd77ea7f64c9","sha256:73130839f4817edd99d49c90254718516dd5d3879d174de7f358ce46cf315a33"],"state_sha256":"5d955a4990f0563618d122e46cb8b56148f8fd938ce8b1a49ae400190d5332cd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YuM23Ehdi+/PY7cuQqEnhOVjZansG09wpx9df+pJzJlRb+v4ogkrmQ2TDXb2G5N7qNA0hG0mfEmNPPLOOIRSAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-29T15:13:32.249086Z","bundle_sha256":"a6bf50fa01d5e3d739e166965b9c18ce05c955ee70579ac4a5e235b995ecc9c5"}}