{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2017:SIJXF5H336U4IKL3PAMLQRPBJV","short_pith_number":"pith:SIJXF5H3","canonical_record":{"source":{"id":"1709.10380","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-29T12:56:25Z","cross_cats_sorted":[],"title_canon_sha256":"093f3e1706202c9c8e35869b387911fddf481ae48deddf560705b77a3c89f955","abstract_canon_sha256":"5d5b88cf105bf39d304ab3e9f5f1e18077fc008d6eea19c77ff2ff5530c0b2ed"},"schema_version":"1.0"},"canonical_sha256":"921372f4fbdfa9c4297b7818b845e14d4d75a697140d1d6db34e90d0478dea2d","source":{"kind":"arxiv","id":"1709.10380","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1709.10380","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"arxiv_version","alias_value":"1709.10380v5","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.10380","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"pith_short_12","alias_value":"SIJXF5H336U4","created_at":"2026-05-18T12:31:43Z"},{"alias_kind":"pith_short_16","alias_value":"SIJXF5H336U4IKL3","created_at":"2026-05-18T12:31:43Z"},{"alias_kind":"pith_short_8","alias_value":"SIJXF5H3","created_at":"2026-05-18T12:31:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2017:SIJXF5H336U4IKL3PAMLQRPBJV","target":"record","payload":{"canonical_record":{"source":{"id":"1709.10380","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-29T12:56:25Z","cross_cats_sorted":[],"title_canon_sha256":"093f3e1706202c9c8e35869b387911fddf481ae48deddf560705b77a3c89f955","abstract_canon_sha256":"5d5b88cf105bf39d304ab3e9f5f1e18077fc008d6eea19c77ff2ff5530c0b2ed"},"schema_version":"1.0"},"canonical_sha256":"921372f4fbdfa9c4297b7818b845e14d4d75a697140d1d6db34e90d0478dea2d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:00:41.438338Z","signature_b64":"fYu8FZz8/XpAqQQhoDH9aKOcDrt9vPSBTTSgdjjE1bNtDnVJP6Mui31yWqRnPRttf+B2WwOp5eFbzRcso+UBDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"921372f4fbdfa9c4297b7818b845e14d4d75a697140d1d6db34e90d0478dea2d","last_reissued_at":"2026-05-18T00:00:41.437936Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:00:41.437936Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1709.10380","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:00:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fO7doySn1eEmfkl0wPrk3UdU30TgE+zecO4AZ+wemVeRHtMoaPGLXcOXX08B9vmm3SWTcsfsyu2sqLUTMeaoCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-24T12:22:41.289467Z"},"content_sha256":"b0520c9ab0ce6258e87557f2cda0b9e2816b565c02dcac61b5adc57883253469","schema_version":"1.0","event_id":"sha256:b0520c9ab0ce6258e87557f2cda0b9e2816b565c02dcac61b5adc57883253469"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2017:SIJXF5H336U4IKL3PAMLQRPBJV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"An Empirical Evaluation of Rule Extraction from Recurrent Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander G. Ororbia II, C. Lee Giles, Kaixuan Zhang, Qinglong Wang, Xinyu Xing, Xue Liu","submitted_at":"2017-09-29T12:56:25Z","abstract_excerpt":"Rule extraction from black-box models is critical in domains that require model validation before implementation, as can be the case in credit scoring and medical diagnosis. Though already a challenging problem in statistical learning in general, the difficulty is even greater when highly non-linear, recursive models, such as recurrent neural networks (RNNs), are fit to data. Here, we study the extraction of rules from second-order recurrent neural networks trained to recognize the Tomita grammars. We show that production rules can be stably extracted from trained RNNs and that in certain case"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1709.10380","kind":"arxiv","version":5},"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":"2026-05-18T00:00:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NHA1mwTJd2YpWbNfn2P8aRDvUia74bpqoa39Scu6ljMLZMtNo0BlxwjDgJVwuImtZ8qB2atkuPqkDnPVGq1DDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-24T12:22:41.290120Z"},"content_sha256":"d748b78dbcd5c073cfd013b6a987e918ac906b7246ba019a64c8d70b892e81db","schema_version":"1.0","event_id":"sha256:d748b78dbcd5c073cfd013b6a987e918ac906b7246ba019a64c8d70b892e81db"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SIJXF5H336U4IKL3PAMLQRPBJV/bundle.json","state_url":"https://pith.science/pith/SIJXF5H336U4IKL3PAMLQRPBJV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SIJXF5H336U4IKL3PAMLQRPBJV/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":"2026-05-24T12:22:41Z","links":{"resolver":"https://pith.science/pith/SIJXF5H336U4IKL3PAMLQRPBJV","bundle":"https://pith.science/pith/SIJXF5H336U4IKL3PAMLQRPBJV/bundle.json","state":"https://pith.science/pith/SIJXF5H336U4IKL3PAMLQRPBJV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SIJXF5H336U4IKL3PAMLQRPBJV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2017:SIJXF5H336U4IKL3PAMLQRPBJV","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":"5d5b88cf105bf39d304ab3e9f5f1e18077fc008d6eea19c77ff2ff5530c0b2ed","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-29T12:56:25Z","title_canon_sha256":"093f3e1706202c9c8e35869b387911fddf481ae48deddf560705b77a3c89f955"},"schema_version":"1.0","source":{"id":"1709.10380","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1709.10380","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"arxiv_version","alias_value":"1709.10380v5","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.10380","created_at":"2026-05-18T00:00:41Z"},{"alias_kind":"pith_short_12","alias_value":"SIJXF5H336U4","created_at":"2026-05-18T12:31:43Z"},{"alias_kind":"pith_short_16","alias_value":"SIJXF5H336U4IKL3","created_at":"2026-05-18T12:31:43Z"},{"alias_kind":"pith_short_8","alias_value":"SIJXF5H3","created_at":"2026-05-18T12:31:43Z"}],"graph_snapshots":[{"event_id":"sha256:d748b78dbcd5c073cfd013b6a987e918ac906b7246ba019a64c8d70b892e81db","target":"graph","created_at":"2026-05-18T00:00:41Z","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":"Rule extraction from black-box models is critical in domains that require model validation before implementation, as can be the case in credit scoring and medical diagnosis. Though already a challenging problem in statistical learning in general, the difficulty is even greater when highly non-linear, recursive models, such as recurrent neural networks (RNNs), are fit to data. Here, we study the extraction of rules from second-order recurrent neural networks trained to recognize the Tomita grammars. We show that production rules can be stably extracted from trained RNNs and that in certain case","authors_text":"Alexander G. Ororbia II, C. Lee Giles, Kaixuan Zhang, Qinglong Wang, Xinyu Xing, Xue Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-29T12:56:25Z","title":"An Empirical Evaluation of Rule Extraction from Recurrent Neural Networks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1709.10380","kind":"arxiv","version":5},"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:b0520c9ab0ce6258e87557f2cda0b9e2816b565c02dcac61b5adc57883253469","target":"record","created_at":"2026-05-18T00:00:41Z","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":"5d5b88cf105bf39d304ab3e9f5f1e18077fc008d6eea19c77ff2ff5530c0b2ed","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-29T12:56:25Z","title_canon_sha256":"093f3e1706202c9c8e35869b387911fddf481ae48deddf560705b77a3c89f955"},"schema_version":"1.0","source":{"id":"1709.10380","kind":"arxiv","version":5}},"canonical_sha256":"921372f4fbdfa9c4297b7818b845e14d4d75a697140d1d6db34e90d0478dea2d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"921372f4fbdfa9c4297b7818b845e14d4d75a697140d1d6db34e90d0478dea2d","first_computed_at":"2026-05-18T00:00:41.437936Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:00:41.437936Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fYu8FZz8/XpAqQQhoDH9aKOcDrt9vPSBTTSgdjjE1bNtDnVJP6Mui31yWqRnPRttf+B2WwOp5eFbzRcso+UBDQ==","signature_status":"signed_v1","signed_at":"2026-05-18T00:00:41.438338Z","signed_message":"canonical_sha256_bytes"},"source_id":"1709.10380","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b0520c9ab0ce6258e87557f2cda0b9e2816b565c02dcac61b5adc57883253469","sha256:d748b78dbcd5c073cfd013b6a987e918ac906b7246ba019a64c8d70b892e81db"],"state_sha256":"6537611b31a8ca05f21daf131b35cb4927c7576cad49323fe2d4492d0aee9ae1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5z7JnSFqq6SQrBWzBDNbSCOviLcKTTeWblp4wYmt2cXAzrPDO3jc3LQoOdaPLFsSXmrHqHtlFVa4r5u5K4plAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-24T12:22:41.293291Z","bundle_sha256":"baa2cbadb6f0f6cbdf0c465f31b326b902446ecc1f092f6e838a19a5654b9c68"}}