{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2013:A2LTIMU54VFK5S5NUGUPF3XVHJ","short_pith_number":"pith:A2LTIMU5","canonical_record":{"source":{"id":"1306.0514","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2013-06-03T17:36:14Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"57a241a36a1a42c3ad4befea975722223ae00342022a991a6f5fecff257c19b3","abstract_canon_sha256":"bd8f59ae0e7498cf4ce1526a72c7e11b8f3e9bf18371a1053f31142be2682194"},"schema_version":"1.0"},"canonical_sha256":"069734329de54aaecbada1a8f2eef53a5f8098e2a6b9c45594970aacb404d8eb","source":{"kind":"arxiv","id":"1306.0514","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1306.0514","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"arxiv_version","alias_value":"1306.0514v4","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1306.0514","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"pith_short_12","alias_value":"A2LTIMU54VFK","created_at":"2026-05-18T12:27:38Z"},{"alias_kind":"pith_short_16","alias_value":"A2LTIMU54VFK5S5N","created_at":"2026-05-18T12:27:38Z"},{"alias_kind":"pith_short_8","alias_value":"A2LTIMU5","created_at":"2026-05-18T12:27:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2013:A2LTIMU54VFK5S5NUGUPF3XVHJ","target":"record","payload":{"canonical_record":{"source":{"id":"1306.0514","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2013-06-03T17:36:14Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"57a241a36a1a42c3ad4befea975722223ae00342022a991a6f5fecff257c19b3","abstract_canon_sha256":"bd8f59ae0e7498cf4ce1526a72c7e11b8f3e9bf18371a1053f31142be2682194"},"schema_version":"1.0"},"canonical_sha256":"069734329de54aaecbada1a8f2eef53a5f8098e2a6b9c45594970aacb404d8eb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:28:05.565856Z","signature_b64":"18T6Vwi2ubn8eQDGEamYqzBq4HfV1viZpdtVwScmD3VMBa9c43OG9uNxJG//FpoanYp1ArIefdiiyzzTb0zaAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"069734329de54aaecbada1a8f2eef53a5f8098e2a6b9c45594970aacb404d8eb","last_reissued_at":"2026-05-18T02:28:05.565416Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:28:05.565416Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1306.0514","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:28:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2Tt7cRv8SQVLql1Lhq8ZK6keUkm/BLPU+k/TCd+tDQVIIJtZgtxHr0V/0DeGbijOcKQ3lJYFGqp9sBEk77YkDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T07:00:29.390369Z"},"content_sha256":"ed9f97667eaaf6a8b396b28cc1b2b7be593843584f095f0525caf0f84ebac04b","schema_version":"1.0","event_id":"sha256:ed9f97667eaaf6a8b396b28cc1b2b7be593843584f095f0525caf0f84ebac04b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2013:A2LTIMU54VFK5S5NUGUPF3XVHJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Riemannian metrics for neural networks II: recurrent networks and learning symbolic data sequences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.NE","authors_text":"Yann Ollivier","submitted_at":"2013-06-03T17:36:14Z","abstract_excerpt":"Recurrent neural networks are powerful models for sequential data, able to represent complex dependencies in the sequence that simpler models such as hidden Markov models cannot handle. Yet they are notoriously hard to train. Here we introduce a training procedure using a gradient ascent in a Riemannian metric: this produces an algorithm independent from design choices such as the encoding of parameters and unit activities. This metric gradient ascent is designed to have an algorithmic cost close to backpropagation through time for sparsely connected networks. We use this procedure on gated le"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1306.0514","kind":"arxiv","version":4},"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":"2026-05-18T02:28:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+NkVN0wG3MCBQFfTmy8MIiioMhRrTyS/eZyosr1vpt0Xd678X6zQ2dNsOYyErUJRPeaitIm9WjB9Gh2+2g6ODw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T07:00:29.391061Z"},"content_sha256":"00a2151908b50cd47b5d8ee054595393446aa0640f74d514bf5100131670d656","schema_version":"1.0","event_id":"sha256:00a2151908b50cd47b5d8ee054595393446aa0640f74d514bf5100131670d656"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/bundle.json","state_url":"https://pith.science/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/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":"2026-06-06T07:00:29Z","links":{"resolver":"https://pith.science/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ","bundle":"https://pith.science/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/bundle.json","state":"https://pith.science/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/A2LTIMU54VFK5S5NUGUPF3XVHJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2013:A2LTIMU54VFK5S5NUGUPF3XVHJ","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":"bd8f59ae0e7498cf4ce1526a72c7e11b8f3e9bf18371a1053f31142be2682194","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2013-06-03T17:36:14Z","title_canon_sha256":"57a241a36a1a42c3ad4befea975722223ae00342022a991a6f5fecff257c19b3"},"schema_version":"1.0","source":{"id":"1306.0514","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1306.0514","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"arxiv_version","alias_value":"1306.0514v4","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1306.0514","created_at":"2026-05-18T02:28:05Z"},{"alias_kind":"pith_short_12","alias_value":"A2LTIMU54VFK","created_at":"2026-05-18T12:27:38Z"},{"alias_kind":"pith_short_16","alias_value":"A2LTIMU54VFK5S5N","created_at":"2026-05-18T12:27:38Z"},{"alias_kind":"pith_short_8","alias_value":"A2LTIMU5","created_at":"2026-05-18T12:27:38Z"}],"graph_snapshots":[{"event_id":"sha256:00a2151908b50cd47b5d8ee054595393446aa0640f74d514bf5100131670d656","target":"graph","created_at":"2026-05-18T02:28:05Z","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 are powerful models for sequential data, able to represent complex dependencies in the sequence that simpler models such as hidden Markov models cannot handle. Yet they are notoriously hard to train. Here we introduce a training procedure using a gradient ascent in a Riemannian metric: this produces an algorithm independent from design choices such as the encoding of parameters and unit activities. This metric gradient ascent is designed to have an algorithmic cost close to backpropagation through time for sparsely connected networks. We use this procedure on gated le","authors_text":"Yann Ollivier","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2013-06-03T17:36:14Z","title":"Riemannian metrics for neural networks II: recurrent networks and learning symbolic data sequences"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1306.0514","kind":"arxiv","version":4},"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:ed9f97667eaaf6a8b396b28cc1b2b7be593843584f095f0525caf0f84ebac04b","target":"record","created_at":"2026-05-18T02:28:05Z","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":"bd8f59ae0e7498cf4ce1526a72c7e11b8f3e9bf18371a1053f31142be2682194","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2013-06-03T17:36:14Z","title_canon_sha256":"57a241a36a1a42c3ad4befea975722223ae00342022a991a6f5fecff257c19b3"},"schema_version":"1.0","source":{"id":"1306.0514","kind":"arxiv","version":4}},"canonical_sha256":"069734329de54aaecbada1a8f2eef53a5f8098e2a6b9c45594970aacb404d8eb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"069734329de54aaecbada1a8f2eef53a5f8098e2a6b9c45594970aacb404d8eb","first_computed_at":"2026-05-18T02:28:05.565416Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:28:05.565416Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"18T6Vwi2ubn8eQDGEamYqzBq4HfV1viZpdtVwScmD3VMBa9c43OG9uNxJG//FpoanYp1ArIefdiiyzzTb0zaAw==","signature_status":"signed_v1","signed_at":"2026-05-18T02:28:05.565856Z","signed_message":"canonical_sha256_bytes"},"source_id":"1306.0514","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ed9f97667eaaf6a8b396b28cc1b2b7be593843584f095f0525caf0f84ebac04b","sha256:00a2151908b50cd47b5d8ee054595393446aa0640f74d514bf5100131670d656"],"state_sha256":"82805558f0a22705a4f74a0952e06dae16322705749b09ec775f4d3772f18d96"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cK+NOMoMBSnxn38AI/8W3IhQrknhe4G47QP0x+hZbMF2IPFFurWy0p7zhPih+/kGmjNPcH2+l3zxkgO0qtdgDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-06T07:00:29.396092Z","bundle_sha256":"590a2250eb849600824e3830e60db5ab78a965b404b5cb514dbab8a0a2ef86b9"}}