{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:WAUY3LLF2FZNXJRBJI4VNVOVSE","short_pith_number":"pith:WAUY3LLF","canonical_record":{"source":{"id":"2306.08756","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-14T21:41:52Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"622c28654d47ea2c902c88b07af106b5298514e8429545f5a6245c6a5fa6cf54","abstract_canon_sha256":"560154e712d53a69f024e7d69d15559e7f70a77d2969396f2e663c5bc868fc47"},"schema_version":"1.0"},"canonical_sha256":"b0298dad65d172dba6214a3956d5d5910fdd7a2338d757632d931bfe2abd7fca","source":{"kind":"arxiv","id":"2306.08756","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.08756","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"arxiv_version","alias_value":"2306.08756v1","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.08756","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_12","alias_value":"WAUY3LLF2FZN","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_16","alias_value":"WAUY3LLF2FZNXJRB","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_8","alias_value":"WAUY3LLF","created_at":"2026-07-05T06:20:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:WAUY3LLF2FZNXJRBJI4VNVOVSE","target":"record","payload":{"canonical_record":{"source":{"id":"2306.08756","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-14T21:41:52Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"622c28654d47ea2c902c88b07af106b5298514e8429545f5a6245c6a5fa6cf54","abstract_canon_sha256":"560154e712d53a69f024e7d69d15559e7f70a77d2969396f2e663c5bc868fc47"},"schema_version":"1.0"},"canonical_sha256":"b0298dad65d172dba6214a3956d5d5910fdd7a2338d757632d931bfe2abd7fca","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:20:56.007010Z","signature_b64":"Gu+Si/1Y4j+hpjo08WbsHVELBNtFuQXqUUzFLDwJqJx6sgWZUl44AtNwakgBFyTyOtseCX28dpZQ6wLLftsuAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b0298dad65d172dba6214a3956d5d5910fdd7a2338d757632d931bfe2abd7fca","last_reissued_at":"2026-07-05T06:20:56.006480Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:20:56.006480Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2306.08756","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":"2026-07-05T06:20:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4gxJCVGAzvq10xIkntg7WReuag/2mH61IEZX2vEOGDQ3d8jiD6UlPDpK4lhsnMZUgbC8kpovACSCRruVdFqcBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:39:16.250932Z"},"content_sha256":"c672d7d5bd172bc549c1d6373b9909ffe91afd80ba7cff46b52d264a90a6cc3d","schema_version":"1.0","event_id":"sha256:c672d7d5bd172bc549c1d6373b9909ffe91afd80ba7cff46b52d264a90a6cc3d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:WAUY3LLF2FZNXJRBJI4VNVOVSE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Recipes for Sequential Pre-training of Multilingual Encoder and Seq2Seq Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Andy Rosenbaum, Anna Rumshisky, Qin Lu, Saleh Soltan, Tobias Falke, Wael Hamza","submitted_at":"2023-06-14T21:41:52Z","abstract_excerpt":"Pre-trained encoder-only and sequence-to-sequence (seq2seq) models each have advantages, however training both model types from scratch is computationally expensive. We explore recipes to improve pre-training efficiency by initializing one model from the other. (1) Extracting the encoder from a seq2seq model, we show it under-performs a Masked Language Modeling (MLM) encoder, particularly on sequence labeling tasks. Variations of masking during seq2seq training, reducing the decoder size, and continuing with a small amount of MLM training do not close the gap. (2) Conversely, using an encoder "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.08756","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2306.08756/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":"2026-07-05T06:20:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2FKWF+2wA4MZoDH1a9qCzYlztK/16jW03Vbv2CcgUvh8AOej2iMQWmoLhlJ6nygh8auibVFX/b20rSplyhJrDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:39:16.251309Z"},"content_sha256":"da5b3e844d85862e37e6d80d8940b302de8e4e6be7ad53b593121cb478c0a789","schema_version":"1.0","event_id":"sha256:da5b3e844d85862e37e6d80d8940b302de8e4e6be7ad53b593121cb478c0a789"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/bundle.json","state_url":"https://pith.science/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/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":"2026-07-07T04:39:16Z","links":{"resolver":"https://pith.science/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE","bundle":"https://pith.science/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/bundle.json","state":"https://pith.science/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WAUY3LLF2FZNXJRBJI4VNVOVSE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:WAUY3LLF2FZNXJRBJI4VNVOVSE","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":"560154e712d53a69f024e7d69d15559e7f70a77d2969396f2e663c5bc868fc47","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-14T21:41:52Z","title_canon_sha256":"622c28654d47ea2c902c88b07af106b5298514e8429545f5a6245c6a5fa6cf54"},"schema_version":"1.0","source":{"id":"2306.08756","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.08756","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"arxiv_version","alias_value":"2306.08756v1","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.08756","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_12","alias_value":"WAUY3LLF2FZN","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_16","alias_value":"WAUY3LLF2FZNXJRB","created_at":"2026-07-05T06:20:56Z"},{"alias_kind":"pith_short_8","alias_value":"WAUY3LLF","created_at":"2026-07-05T06:20:56Z"}],"graph_snapshots":[{"event_id":"sha256:da5b3e844d85862e37e6d80d8940b302de8e4e6be7ad53b593121cb478c0a789","target":"graph","created_at":"2026-07-05T06:20:56Z","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/2306.08756/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pre-trained encoder-only and sequence-to-sequence (seq2seq) models each have advantages, however training both model types from scratch is computationally expensive. We explore recipes to improve pre-training efficiency by initializing one model from the other. (1) Extracting the encoder from a seq2seq model, we show it under-performs a Masked Language Modeling (MLM) encoder, particularly on sequence labeling tasks. Variations of masking during seq2seq training, reducing the decoder size, and continuing with a small amount of MLM training do not close the gap. (2) Conversely, using an encoder ","authors_text":"Andy Rosenbaum, Anna Rumshisky, Qin Lu, Saleh Soltan, Tobias Falke, Wael Hamza","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-14T21:41:52Z","title":"Recipes for Sequential Pre-training of Multilingual Encoder and Seq2Seq Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.08756","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:c672d7d5bd172bc549c1d6373b9909ffe91afd80ba7cff46b52d264a90a6cc3d","target":"record","created_at":"2026-07-05T06:20:56Z","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":"560154e712d53a69f024e7d69d15559e7f70a77d2969396f2e663c5bc868fc47","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-14T21:41:52Z","title_canon_sha256":"622c28654d47ea2c902c88b07af106b5298514e8429545f5a6245c6a5fa6cf54"},"schema_version":"1.0","source":{"id":"2306.08756","kind":"arxiv","version":1}},"canonical_sha256":"b0298dad65d172dba6214a3956d5d5910fdd7a2338d757632d931bfe2abd7fca","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b0298dad65d172dba6214a3956d5d5910fdd7a2338d757632d931bfe2abd7fca","first_computed_at":"2026-07-05T06:20:56.006480Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:20:56.006480Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Gu+Si/1Y4j+hpjo08WbsHVELBNtFuQXqUUzFLDwJqJx6sgWZUl44AtNwakgBFyTyOtseCX28dpZQ6wLLftsuAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T06:20:56.007010Z","signed_message":"canonical_sha256_bytes"},"source_id":"2306.08756","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c672d7d5bd172bc549c1d6373b9909ffe91afd80ba7cff46b52d264a90a6cc3d","sha256:da5b3e844d85862e37e6d80d8940b302de8e4e6be7ad53b593121cb478c0a789"],"state_sha256":"fc6be747814d8d9e0fd8a5c059aba2cc52d028d58a14d4f7956617e52b1a9db1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LmAK5SqtcBRJ4ZrDDSeFFG29lIHiUkNpXvBl7VxX8ELfPzq/ibpp4H2ijMWCTXsCe0qKvv07ov0uS0nMHd5HAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T04:39:16.253479Z","bundle_sha256":"948f16a4a2af8aa77476025d38bb6c82602f60c5f27a960f49fab6a7a16e2eba"}}