{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:RGM6QJFNQCKUV42KGZJW34RL2J","short_pith_number":"pith:RGM6QJFN","canonical_record":{"source":{"id":"2203.13151","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-24T16:12:21Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"8b4053d216862e25ff6e87669b60dbff4b27839cca76be43d4298ffada7de29d","abstract_canon_sha256":"9d782ba7190fd9e36b6040eba156302dc9570d5f612d38479a1b3ce8ffbb756d"},"schema_version":"1.0"},"canonical_sha256":"8999e824ad80954af34a36536df22bd25ea85b9c8c9009d045103d5293570c54","source":{"kind":"arxiv","id":"2203.13151","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2203.13151","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"arxiv_version","alias_value":"2203.13151v2","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.13151","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_12","alias_value":"RGM6QJFNQCKU","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_16","alias_value":"RGM6QJFNQCKUV42K","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_8","alias_value":"RGM6QJFN","created_at":"2026-07-05T06:15:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:RGM6QJFNQCKUV42KGZJW34RL2J","target":"record","payload":{"canonical_record":{"source":{"id":"2203.13151","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-24T16:12:21Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"8b4053d216862e25ff6e87669b60dbff4b27839cca76be43d4298ffada7de29d","abstract_canon_sha256":"9d782ba7190fd9e36b6040eba156302dc9570d5f612d38479a1b3ce8ffbb756d"},"schema_version":"1.0"},"canonical_sha256":"8999e824ad80954af34a36536df22bd25ea85b9c8c9009d045103d5293570c54","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:15:12.602527Z","signature_b64":"02gF6QsYEmfyQR9MyVLAlTZEr3b2tuminBLHHRzL4ZRRYIa81yYXV/k/SU91b4Dvhr3cQt5hSO0pgWEuviByCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8999e824ad80954af34a36536df22bd25ea85b9c8c9009d045103d5293570c54","last_reissued_at":"2026-07-05T06:15:12.602025Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:15:12.602025Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2203.13151","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T06:15:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UKVsFwFD6r4VKE/iRIGMRb5I24XSvDyLU9CGIZw1a9HxpLEr5qy30Rf3MX7na9jRLQiyoNK1/9mt88xJf1BhAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:23:54.298349Z"},"content_sha256":"b63fccdffe76ac8fce84e35a3ae111ef738b8b1a9eb39f5c7e4677da50ec4b70","schema_version":"1.0","event_id":"sha256:b63fccdffe76ac8fce84e35a3ae111ef738b8b1a9eb39f5c7e4677da50ec4b70"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:RGM6QJFNQCKUV42KGZJW34RL2J","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Multi-armed bandits for resource efficient, online optimization of language model pre-training: the use case of dynamic masking","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"I\\~nigo Urteaga, Moulay-Za\\\"idane Dra\\\"idia, Shahram Khadivi, Tomer Lancewicki","submitted_at":"2022-03-24T16:12:21Z","abstract_excerpt":"We design and evaluate a Bayesian optimization framework for resource efficient pre-training of Transformer-based language models (TLMs). TLM pre-training requires high computational resources and introduces many unresolved design choices, such as selecting its pre-training hyperparameters. We propose a multi-armed bandit framework for the sequential selection of TLM pre-training hyperparameters, aimed at optimizing language model performance, in a resource efficient manner. We design a Thompson sampling algorithm, with a surrogate Gaussian process reward model of the Masked Language Model (ML"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.13151","kind":"arxiv","version":2},"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/2203.13151/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":"2026-07-05T06:15:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NBJ78elJhvq0pTJL9VbYgxxoisuKOEK9IKOjQGNsLC2D9hPRwLszCkBkhF1FXe+Ro0NENSXTRPI6jjsE6YnbDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:23:54.298877Z"},"content_sha256":"40133d94da5ea576d9aea1f789073bed2256a2d31ac81117d01cbe9a2733733f","schema_version":"1.0","event_id":"sha256:40133d94da5ea576d9aea1f789073bed2256a2d31ac81117d01cbe9a2733733f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RGM6QJFNQCKUV42KGZJW34RL2J/bundle.json","state_url":"https://pith.science/pith/RGM6QJFNQCKUV42KGZJW34RL2J/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RGM6QJFNQCKUV42KGZJW34RL2J/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":"2026-08-05T02:23:54Z","links":{"resolver":"https://pith.science/pith/RGM6QJFNQCKUV42KGZJW34RL2J","bundle":"https://pith.science/pith/RGM6QJFNQCKUV42KGZJW34RL2J/bundle.json","state":"https://pith.science/pith/RGM6QJFNQCKUV42KGZJW34RL2J/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RGM6QJFNQCKUV42KGZJW34RL2J/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:RGM6QJFNQCKUV42KGZJW34RL2J","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":"9d782ba7190fd9e36b6040eba156302dc9570d5f612d38479a1b3ce8ffbb756d","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-24T16:12:21Z","title_canon_sha256":"8b4053d216862e25ff6e87669b60dbff4b27839cca76be43d4298ffada7de29d"},"schema_version":"1.0","source":{"id":"2203.13151","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2203.13151","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"arxiv_version","alias_value":"2203.13151v2","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.13151","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_12","alias_value":"RGM6QJFNQCKU","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_16","alias_value":"RGM6QJFNQCKUV42K","created_at":"2026-07-05T06:15:12Z"},{"alias_kind":"pith_short_8","alias_value":"RGM6QJFN","created_at":"2026-07-05T06:15:12Z"}],"graph_snapshots":[{"event_id":"sha256:40133d94da5ea576d9aea1f789073bed2256a2d31ac81117d01cbe9a2733733f","target":"graph","created_at":"2026-07-05T06:15:12Z","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/2203.13151/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We design and evaluate a Bayesian optimization framework for resource efficient pre-training of Transformer-based language models (TLMs). TLM pre-training requires high computational resources and introduces many unresolved design choices, such as selecting its pre-training hyperparameters. We propose a multi-armed bandit framework for the sequential selection of TLM pre-training hyperparameters, aimed at optimizing language model performance, in a resource efficient manner. We design a Thompson sampling algorithm, with a surrogate Gaussian process reward model of the Masked Language Model (ML","authors_text":"I\\~nigo Urteaga, Moulay-Za\\\"idane Dra\\\"idia, Shahram Khadivi, Tomer Lancewicki","cross_cats":["cs.LG","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-24T16:12:21Z","title":"Multi-armed bandits for resource efficient, online optimization of language model pre-training: the use case of dynamic masking"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.13151","kind":"arxiv","version":2},"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:b63fccdffe76ac8fce84e35a3ae111ef738b8b1a9eb39f5c7e4677da50ec4b70","target":"record","created_at":"2026-07-05T06:15:12Z","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":"9d782ba7190fd9e36b6040eba156302dc9570d5f612d38479a1b3ce8ffbb756d","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-24T16:12:21Z","title_canon_sha256":"8b4053d216862e25ff6e87669b60dbff4b27839cca76be43d4298ffada7de29d"},"schema_version":"1.0","source":{"id":"2203.13151","kind":"arxiv","version":2}},"canonical_sha256":"8999e824ad80954af34a36536df22bd25ea85b9c8c9009d045103d5293570c54","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8999e824ad80954af34a36536df22bd25ea85b9c8c9009d045103d5293570c54","first_computed_at":"2026-07-05T06:15:12.602025Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:15:12.602025Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"02gF6QsYEmfyQR9MyVLAlTZEr3b2tuminBLHHRzL4ZRRYIa81yYXV/k/SU91b4Dvhr3cQt5hSO0pgWEuviByCA==","signature_status":"signed_v1","signed_at":"2026-07-05T06:15:12.602527Z","signed_message":"canonical_sha256_bytes"},"source_id":"2203.13151","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b63fccdffe76ac8fce84e35a3ae111ef738b8b1a9eb39f5c7e4677da50ec4b70","sha256:40133d94da5ea576d9aea1f789073bed2256a2d31ac81117d01cbe9a2733733f"],"state_sha256":"0f808f058eb71de4672634ed9999e64973c5bd28119f1e1a56c5ccacedcd8583"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oHfP3qwi4rokOQ8fSalSxjOi/nCiEyNRgW6ErPZG9lsK8WodCridmK6fzAxFIjkrrGrVJgIyogyk2oCQYrIpBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T02:23:54.304500Z","bundle_sha256":"858651bc113425bb01762e03da3a4dc1daf051acacf82f723044917fc1425290"}}