{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:UJC2C2IKNZHQ3PELBXIHGFQUO7","short_pith_number":"pith:UJC2C2IK","canonical_record":{"source":{"id":"2310.00796","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-01T21:19:12Z","cross_cats_sorted":[],"title_canon_sha256":"ddb58d89acaa6f9cdc52b9813f7fecb0e19486f7e8739107bdc4735d2508933f","abstract_canon_sha256":"ecf0de6b2b2e882625f0a634094436858e6408e1039be74534abfe017bd4e111"},"schema_version":"1.0"},"canonical_sha256":"a245a1690a6e4f0dbc8b0dd073161477fad25b227047454631181aa21458e95e","source":{"kind":"arxiv","id":"2310.00796","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.00796","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"arxiv_version","alias_value":"2310.00796v3","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.00796","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_12","alias_value":"UJC2C2IKNZHQ","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_16","alias_value":"UJC2C2IKNZHQ3PEL","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_8","alias_value":"UJC2C2IK","created_at":"2026-07-05T08:42:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:UJC2C2IKNZHQ3PELBXIHGFQUO7","target":"record","payload":{"canonical_record":{"source":{"id":"2310.00796","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-01T21:19:12Z","cross_cats_sorted":[],"title_canon_sha256":"ddb58d89acaa6f9cdc52b9813f7fecb0e19486f7e8739107bdc4735d2508933f","abstract_canon_sha256":"ecf0de6b2b2e882625f0a634094436858e6408e1039be74534abfe017bd4e111"},"schema_version":"1.0"},"canonical_sha256":"a245a1690a6e4f0dbc8b0dd073161477fad25b227047454631181aa21458e95e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:42:21.181543Z","signature_b64":"YKhym9kOXdbohyjF+F/YQMzDy5SZt+nnSqc/ymILrpQGy6y+mLdf1K/PM1CKWqzkXDl50JZnyONw6FHcLRGhCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a245a1690a6e4f0dbc8b0dd073161477fad25b227047454631181aa21458e95e","last_reissued_at":"2026-07-05T08:42:21.181112Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:42:21.181112Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.00796","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:42:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SAk38oDUxMlzEmX5A5H/S6UkhHZ1PAaZL/pHf3DXh5LIYXOP1Xz9HS18S2dS53P46c82N4chGkiaspUD6cRXBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:45:36.950410Z"},"content_sha256":"3a31deda81ce840e165cdfda2fcb079307c1ec924476fa7fc1c2d494fbe0d110","schema_version":"1.0","event_id":"sha256:3a31deda81ce840e165cdfda2fcb079307c1ec924476fa7fc1c2d494fbe0d110"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:UJC2C2IKNZHQ3PELBXIHGFQUO7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SIP: Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Koller, Ivan Titov, Matthias Lindemann","submitted_at":"2023-10-01T21:19:12Z","abstract_excerpt":"Strong inductive biases enable learning from little data and help generalization outside of the training distribution. Popular neural architectures such as Transformers lack strong structural inductive biases for seq2seq NLP tasks on their own. Consequently, they struggle with systematic generalization beyond the training distribution, e.g. with extrapolating to longer inputs, even when pre-trained on large amounts of text. We show how a structural inductive bias can be efficiently injected into a seq2seq model by pre-training it to simulate structural transformations on synthetic data. Specif"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.00796","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/2310.00796/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:42:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cA+n1YzveLDpJ4sSpGSKzdoe0ehkKIgc+Wg+tNz7TI/qAT7i69OVeZ3kl8FdeAbh1tgRM/T9gK5PU1jiZkcPAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:45:36.951161Z"},"content_sha256":"5bdd5f978f40f2a37542fd8a236e7152c0b46f1a99aa6619de7eb2707c3ce6fa","schema_version":"1.0","event_id":"sha256:5bdd5f978f40f2a37542fd8a236e7152c0b46f1a99aa6619de7eb2707c3ce6fa"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/bundle.json","state_url":"https://pith.science/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/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":"2026-08-10T08:45:36Z","links":{"resolver":"https://pith.science/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7","bundle":"https://pith.science/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/bundle.json","state":"https://pith.science/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UJC2C2IKNZHQ3PELBXIHGFQUO7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:UJC2C2IKNZHQ3PELBXIHGFQUO7","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":"ecf0de6b2b2e882625f0a634094436858e6408e1039be74534abfe017bd4e111","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-01T21:19:12Z","title_canon_sha256":"ddb58d89acaa6f9cdc52b9813f7fecb0e19486f7e8739107bdc4735d2508933f"},"schema_version":"1.0","source":{"id":"2310.00796","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.00796","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"arxiv_version","alias_value":"2310.00796v3","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.00796","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_12","alias_value":"UJC2C2IKNZHQ","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_16","alias_value":"UJC2C2IKNZHQ3PEL","created_at":"2026-07-05T08:42:21Z"},{"alias_kind":"pith_short_8","alias_value":"UJC2C2IK","created_at":"2026-07-05T08:42:21Z"}],"graph_snapshots":[{"event_id":"sha256:5bdd5f978f40f2a37542fd8a236e7152c0b46f1a99aa6619de7eb2707c3ce6fa","target":"graph","created_at":"2026-07-05T08:42:21Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2310.00796/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Strong inductive biases enable learning from little data and help generalization outside of the training distribution. Popular neural architectures such as Transformers lack strong structural inductive biases for seq2seq NLP tasks on their own. Consequently, they struggle with systematic generalization beyond the training distribution, e.g. with extrapolating to longer inputs, even when pre-trained on large amounts of text. We show how a structural inductive bias can be efficiently injected into a seq2seq model by pre-training it to simulate structural transformations on synthetic data. Specif","authors_text":"Alexander Koller, Ivan Titov, Matthias Lindemann","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-01T21:19:12Z","title":"SIP: Injecting a Structural Inductive Bias into a Seq2Seq Model by Simulation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.00796","kind":"arxiv","version":3},"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":"sha256:3a31deda81ce840e165cdfda2fcb079307c1ec924476fa7fc1c2d494fbe0d110","target":"record","created_at":"2026-07-05T08:42:21Z","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":"ecf0de6b2b2e882625f0a634094436858e6408e1039be74534abfe017bd4e111","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-01T21:19:12Z","title_canon_sha256":"ddb58d89acaa6f9cdc52b9813f7fecb0e19486f7e8739107bdc4735d2508933f"},"schema_version":"1.0","source":{"id":"2310.00796","kind":"arxiv","version":3}},"canonical_sha256":"a245a1690a6e4f0dbc8b0dd073161477fad25b227047454631181aa21458e95e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a245a1690a6e4f0dbc8b0dd073161477fad25b227047454631181aa21458e95e","first_computed_at":"2026-07-05T08:42:21.181112Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:42:21.181112Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"YKhym9kOXdbohyjF+F/YQMzDy5SZt+nnSqc/ymILrpQGy6y+mLdf1K/PM1CKWqzkXDl50JZnyONw6FHcLRGhCA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:42:21.181543Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.00796","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3a31deda81ce840e165cdfda2fcb079307c1ec924476fa7fc1c2d494fbe0d110","sha256:5bdd5f978f40f2a37542fd8a236e7152c0b46f1a99aa6619de7eb2707c3ce6fa"],"state_sha256":"62e06dea566c906ba45fbace286d89127c5a26ceddc457ba6a37a3e70b8e0680"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yH4ViQyfeEbGFlIN/MMH/5J1pIMlhj8krb18ic2wSpPKnC3a8E4VSpbEKfD7JNS0rVYf8iD3UlnhiXQPYPbiBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T08:45:36.959590Z","bundle_sha256":"f9b20e7697c017be284d885a864bca54919ac288f05c6cfe23c2689a0e6c74b0"}}