{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:6SCBZOKQIDE2HJA7BVDF5WDJHT","short_pith_number":"pith:6SCBZOKQ","canonical_record":{"source":{"id":"1904.04163","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-04-08T16:16:01Z","cross_cats_sorted":[],"title_canon_sha256":"1d985dc54e08f9bf78c2e143c24e5ef6fe53f4cc0221956abf1c6d0749c49899","abstract_canon_sha256":"1d5c48cadce5dc56714dc35418014aa6d79ee9349267ff4f05ab560b3ae23444"},"schema_version":"1.0"},"canonical_sha256":"f4841cb95040c9a3a41f0d465ed8693cc5a769a100d792a1e6179c71a266997f","source":{"kind":"arxiv","id":"1904.04163","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1904.04163","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"arxiv_version","alias_value":"1904.04163v1","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.04163","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"pith_short_12","alias_value":"6SCBZOKQIDE2","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_16","alias_value":"6SCBZOKQIDE2HJA7","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_8","alias_value":"6SCBZOKQ","created_at":"2026-05-18T12:33:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:6SCBZOKQIDE2HJA7BVDF5WDJHT","target":"record","payload":{"canonical_record":{"source":{"id":"1904.04163","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-04-08T16:16:01Z","cross_cats_sorted":[],"title_canon_sha256":"1d985dc54e08f9bf78c2e143c24e5ef6fe53f4cc0221956abf1c6d0749c49899","abstract_canon_sha256":"1d5c48cadce5dc56714dc35418014aa6d79ee9349267ff4f05ab560b3ae23444"},"schema_version":"1.0"},"canonical_sha256":"f4841cb95040c9a3a41f0d465ed8693cc5a769a100d792a1e6179c71a266997f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:49:06.891804Z","signature_b64":"mTEX5wrjuJHlclsiXxCKR4JE67McJ08ffpiyn+6XYYoLS2xpTRtpY8EnVkeagL0iS+Xdv4zAPEkF0+d4NuY+AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f4841cb95040c9a3a41f0d465ed8693cc5a769a100d792a1e6179c71a266997f","last_reissued_at":"2026-05-17T23:49:06.891090Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:49:06.891090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1904.04163","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:49:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BxiosmABbD4QENpqabA/Ftq1CJGcef9otMsjrFoPuGcJhSeyLk/MS18Dcg5dEe2B3/CAqfeHf1rrQ1nOJK0BBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T04:30:21.993447Z"},"content_sha256":"3f4e2642a2df4810f0a51557e1badba4c588874741166199ab2bcdc70431034c","schema_version":"1.0","event_id":"sha256:3f4e2642a2df4810f0a51557e1badba4c588874741166199ab2bcdc70431034c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:6SCBZOKQIDE2HJA7BVDF5WDJHT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Knowledge Distillation For Recurrent Neural Network Language Modeling With Trust Regularization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoyu Sheng, Mei-Yuh Hwang, Xin Lei, Yangyang Shi","submitted_at":"2019-04-08T16:16:01Z","abstract_excerpt":"Recurrent Neural Networks (RNNs) have dominated language modeling because of their superior performance over traditional N-gram based models. In many applications, a large Recurrent Neural Network language model (RNNLM) or an ensemble of several RNNLMs is used. These models have large memory footprints and require heavy computation. In this paper, we examine the effect of applying knowledge distillation in reducing the model size for RNNLMs. In addition, we propose a trust regularization method to improve the knowledge distillation training for RNNLMs. Using knowledge distillation with trust r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.04163","kind":"arxiv","version":1},"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":"2026-05-17T23:49:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E5JWcWLqtRbZB9JDhbK6m4386y8w8OxORXhEDLAhFFgUY84Q2Vk97rgLBO59TDL4x4iYyjXF/5GlsNkwAsaZAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T04:30:21.993796Z"},"content_sha256":"88045b56e005f65dda29a2d3b73121221fbff5efebdce8cee7a5eb29962aab72","schema_version":"1.0","event_id":"sha256:88045b56e005f65dda29a2d3b73121221fbff5efebdce8cee7a5eb29962aab72"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/bundle.json","state_url":"https://pith.science/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/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":"2026-06-05T04:30:21Z","links":{"resolver":"https://pith.science/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT","bundle":"https://pith.science/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/bundle.json","state":"https://pith.science/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6SCBZOKQIDE2HJA7BVDF5WDJHT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:6SCBZOKQIDE2HJA7BVDF5WDJHT","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":"1d5c48cadce5dc56714dc35418014aa6d79ee9349267ff4f05ab560b3ae23444","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-04-08T16:16:01Z","title_canon_sha256":"1d985dc54e08f9bf78c2e143c24e5ef6fe53f4cc0221956abf1c6d0749c49899"},"schema_version":"1.0","source":{"id":"1904.04163","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1904.04163","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"arxiv_version","alias_value":"1904.04163v1","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.04163","created_at":"2026-05-17T23:49:06Z"},{"alias_kind":"pith_short_12","alias_value":"6SCBZOKQIDE2","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_16","alias_value":"6SCBZOKQIDE2HJA7","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_8","alias_value":"6SCBZOKQ","created_at":"2026-05-18T12:33:10Z"}],"graph_snapshots":[{"event_id":"sha256:88045b56e005f65dda29a2d3b73121221fbff5efebdce8cee7a5eb29962aab72","target":"graph","created_at":"2026-05-17T23:49:06Z","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":"Recurrent Neural Networks (RNNs) have dominated language modeling because of their superior performance over traditional N-gram based models. In many applications, a large Recurrent Neural Network language model (RNNLM) or an ensemble of several RNNLMs is used. These models have large memory footprints and require heavy computation. In this paper, we examine the effect of applying knowledge distillation in reducing the model size for RNNLMs. In addition, we propose a trust regularization method to improve the knowledge distillation training for RNNLMs. Using knowledge distillation with trust r","authors_text":"Haoyu Sheng, Mei-Yuh Hwang, Xin Lei, Yangyang Shi","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-04-08T16:16:01Z","title":"Knowledge Distillation For Recurrent Neural Network Language Modeling With Trust Regularization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.04163","kind":"arxiv","version":1},"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":"sha256:3f4e2642a2df4810f0a51557e1badba4c588874741166199ab2bcdc70431034c","target":"record","created_at":"2026-05-17T23:49:06Z","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":"1d5c48cadce5dc56714dc35418014aa6d79ee9349267ff4f05ab560b3ae23444","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-04-08T16:16:01Z","title_canon_sha256":"1d985dc54e08f9bf78c2e143c24e5ef6fe53f4cc0221956abf1c6d0749c49899"},"schema_version":"1.0","source":{"id":"1904.04163","kind":"arxiv","version":1}},"canonical_sha256":"f4841cb95040c9a3a41f0d465ed8693cc5a769a100d792a1e6179c71a266997f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f4841cb95040c9a3a41f0d465ed8693cc5a769a100d792a1e6179c71a266997f","first_computed_at":"2026-05-17T23:49:06.891090Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:49:06.891090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mTEX5wrjuJHlclsiXxCKR4JE67McJ08ffpiyn+6XYYoLS2xpTRtpY8EnVkeagL0iS+Xdv4zAPEkF0+d4NuY+AQ==","signature_status":"signed_v1","signed_at":"2026-05-17T23:49:06.891804Z","signed_message":"canonical_sha256_bytes"},"source_id":"1904.04163","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3f4e2642a2df4810f0a51557e1badba4c588874741166199ab2bcdc70431034c","sha256:88045b56e005f65dda29a2d3b73121221fbff5efebdce8cee7a5eb29962aab72"],"state_sha256":"cd7fa49c0018a53adb7ae3b66fcb3f66e4fc858196f02721f6219d848a2b7667"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GveeEdjmg/osWb/0rlXiyVEGyawVhiZqP9FH2StgAlDbLj7Fpt+WbKR1EhfYXrT+vl828VsK0ZPB2orMzrMCAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T04:30:21.996167Z","bundle_sha256":"e2a858d278b36fb8ce6360a87d52414aa706e54848b616b952231d7c1509f18d"}}