{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:KF4FLEHP5OWBHZEK2JYRVJNWN2","short_pith_number":"pith:KF4FLEHP","canonical_record":{"source":{"id":"2410.03735","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-30T20:49:54Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3fdeccd5a7d937677c9f2a87759cf3dd31fbae797b17af4d3bcacd1ce1e9c1ea","abstract_canon_sha256":"b6a3181321bb841d1cb86d9f034a0b15a3e1ccd871c7717817fe1accef8b9080"},"schema_version":"1.0"},"canonical_sha256":"51785590efebac13e48ad2711aa5b66ea4d49dc23a236d2af27a3d4b63461262","source":{"kind":"arxiv","id":"2410.03735","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.03735","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"arxiv_version","alias_value":"2410.03735v2","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03735","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_12","alias_value":"KF4FLEHP5OWB","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_16","alias_value":"KF4FLEHP5OWBHZEK","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_8","alias_value":"KF4FLEHP","created_at":"2026-07-05T10:28:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:KF4FLEHP5OWBHZEK2JYRVJNWN2","target":"record","payload":{"canonical_record":{"source":{"id":"2410.03735","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-30T20:49:54Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3fdeccd5a7d937677c9f2a87759cf3dd31fbae797b17af4d3bcacd1ce1e9c1ea","abstract_canon_sha256":"b6a3181321bb841d1cb86d9f034a0b15a3e1ccd871c7717817fe1accef8b9080"},"schema_version":"1.0"},"canonical_sha256":"51785590efebac13e48ad2711aa5b66ea4d49dc23a236d2af27a3d4b63461262","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:28:23.566853Z","signature_b64":"Vsu3rGHFjdM6Pj6hXjvCPr9rrFS3XGxKKai0L3Pk/miDaD0avtI8Dw2TZUXPM6kF3L+O76ODcBsPrQR97lzIDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51785590efebac13e48ad2711aa5b66ea4d49dc23a236d2af27a3d4b63461262","last_reissued_at":"2026-07-05T10:28:23.566145Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:28:23.566145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.03735","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":"2026-07-05T10:28:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6LExLkd7MbcOZgsiyrh/ef38xpAgGlOXHMa8FXmzq9JeSzLs0iSywLaaaZ+68LbnSzYUESfI8ARKvX9UVlJSCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:27:27.284424Z"},"content_sha256":"6644a8f037b2534894f9fdd31ebd246e6b7ab0b2ff1930aeafc704e530632977","schema_version":"1.0","event_id":"sha256:6644a8f037b2534894f9fdd31ebd246e6b7ab0b2ff1930aeafc704e530632977"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:KF4FLEHP5OWBHZEK2JYRVJNWN2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Task-Adaptive Pretrained Language Models via Clustered-Importance Sampling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"David Grangier, Pierre Ablin, Simin Fan, Skyler Seto","submitted_at":"2024-09-30T20:49:54Z","abstract_excerpt":"Specialist language models (LMs) focus on a specific task or domain on which they often outperform generalist LMs of the same size. However, the specialist data needed to pretrain these models is only available in limited amount for most tasks. In this work, we build specialist models from large generalist training sets instead. We propose a novel method, ClusteRed Importance SamPling (CRISP). CRISP clusters the generalist dataset and samples from these clusters based on their frequencies in the smaller specialist dataset. It is scalable, suitable for both pretraining and continued pretraining"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03735","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/2410.03735/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":"2026-07-05T10:28:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Hbe4NOaSGaOI4gD/uGen/AdsXs+5fmPzrhwdF5vI6zhJ8NBkDCeomBzhmwTl/RCOu8CDxCqOvrIQiSxQ6QJEDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:27:27.284925Z"},"content_sha256":"0736bab9b8d9ab04429583e80d93be496c4b53062617df81c608faa03c1106ca","schema_version":"1.0","event_id":"sha256:0736bab9b8d9ab04429583e80d93be496c4b53062617df81c608faa03c1106ca"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/bundle.json","state_url":"https://pith.science/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/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":"2026-08-07T21:27:27Z","links":{"resolver":"https://pith.science/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2","bundle":"https://pith.science/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/bundle.json","state":"https://pith.science/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KF4FLEHP5OWBHZEK2JYRVJNWN2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:KF4FLEHP5OWBHZEK2JYRVJNWN2","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":"b6a3181321bb841d1cb86d9f034a0b15a3e1ccd871c7717817fe1accef8b9080","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-30T20:49:54Z","title_canon_sha256":"3fdeccd5a7d937677c9f2a87759cf3dd31fbae797b17af4d3bcacd1ce1e9c1ea"},"schema_version":"1.0","source":{"id":"2410.03735","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.03735","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"arxiv_version","alias_value":"2410.03735v2","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03735","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_12","alias_value":"KF4FLEHP5OWB","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_16","alias_value":"KF4FLEHP5OWBHZEK","created_at":"2026-07-05T10:28:23Z"},{"alias_kind":"pith_short_8","alias_value":"KF4FLEHP","created_at":"2026-07-05T10:28:23Z"}],"graph_snapshots":[{"event_id":"sha256:0736bab9b8d9ab04429583e80d93be496c4b53062617df81c608faa03c1106ca","target":"graph","created_at":"2026-07-05T10:28:23Z","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/2410.03735/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Specialist language models (LMs) focus on a specific task or domain on which they often outperform generalist LMs of the same size. However, the specialist data needed to pretrain these models is only available in limited amount for most tasks. In this work, we build specialist models from large generalist training sets instead. We propose a novel method, ClusteRed Importance SamPling (CRISP). CRISP clusters the generalist dataset and samples from these clusters based on their frequencies in the smaller specialist dataset. It is scalable, suitable for both pretraining and continued pretraining","authors_text":"David Grangier, Pierre Ablin, Simin Fan, Skyler Seto","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-30T20:49:54Z","title":"Task-Adaptive Pretrained Language Models via Clustered-Importance Sampling"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03735","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":"sha256:6644a8f037b2534894f9fdd31ebd246e6b7ab0b2ff1930aeafc704e530632977","target":"record","created_at":"2026-07-05T10:28:23Z","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":"b6a3181321bb841d1cb86d9f034a0b15a3e1ccd871c7717817fe1accef8b9080","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-30T20:49:54Z","title_canon_sha256":"3fdeccd5a7d937677c9f2a87759cf3dd31fbae797b17af4d3bcacd1ce1e9c1ea"},"schema_version":"1.0","source":{"id":"2410.03735","kind":"arxiv","version":2}},"canonical_sha256":"51785590efebac13e48ad2711aa5b66ea4d49dc23a236d2af27a3d4b63461262","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"51785590efebac13e48ad2711aa5b66ea4d49dc23a236d2af27a3d4b63461262","first_computed_at":"2026-07-05T10:28:23.566145Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:28:23.566145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Vsu3rGHFjdM6Pj6hXjvCPr9rrFS3XGxKKai0L3Pk/miDaD0avtI8Dw2TZUXPM6kF3L+O76ODcBsPrQR97lzIDw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:28:23.566853Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.03735","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6644a8f037b2534894f9fdd31ebd246e6b7ab0b2ff1930aeafc704e530632977","sha256:0736bab9b8d9ab04429583e80d93be496c4b53062617df81c608faa03c1106ca"],"state_sha256":"372ecb5e72c2dbba328ab8fab1c2cd7ba6b2d0e1ae6fce7cbefe093c5f1d35af"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gUwndgMWGJpLgUsV8SnDuxwEncXMjzluOb/cYNWcMaGnTCBhC21k5QpjhIQyBSo1PyLS1Q6i/Fr0Z3oObz/ZDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T21:27:27.288706Z","bundle_sha256":"c9c0d11f7342aa0b3d80c69be35444276a0300d1ab19cb9d7bb6a25aac5c01ac"}}