{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:UVKKH7MFGHQG22B7POPDPATBPP","short_pith_number":"pith:UVKKH7MF","canonical_record":{"source":{"id":"2402.14270","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-22T04:10:57Z","cross_cats_sorted":[],"title_canon_sha256":"f95bb28957cc0cb08b299f7ea0c42c81321e759bf7442524fa1927f3688dc637","abstract_canon_sha256":"f9b02e508659b8a20edf7a162d484037f4a51d6e18a872b75b4cc46bdd7a2b90"},"schema_version":"1.0"},"canonical_sha256":"a554a3fd8531e06d683f7b9e3782617bf4a3ae086c52643a69fbcd8c0b697933","source":{"kind":"arxiv","id":"2402.14270","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.14270","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"arxiv_version","alias_value":"2402.14270v2","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.14270","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_12","alias_value":"UVKKH7MFGHQG","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_16","alias_value":"UVKKH7MFGHQG22B7","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_8","alias_value":"UVKKH7MF","created_at":"2026-07-05T07:50:55Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:UVKKH7MFGHQG22B7POPDPATBPP","target":"record","payload":{"canonical_record":{"source":{"id":"2402.14270","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-22T04:10:57Z","cross_cats_sorted":[],"title_canon_sha256":"f95bb28957cc0cb08b299f7ea0c42c81321e759bf7442524fa1927f3688dc637","abstract_canon_sha256":"f9b02e508659b8a20edf7a162d484037f4a51d6e18a872b75b4cc46bdd7a2b90"},"schema_version":"1.0"},"canonical_sha256":"a554a3fd8531e06d683f7b9e3782617bf4a3ae086c52643a69fbcd8c0b697933","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:50:55.848390Z","signature_b64":"6dRnpqOEJl/GnPySrH3Vs+0NuRt2piNfr6ytVs6aabak0Yk+D6VOJC/P+ejfufe8dWvxiOjCc4593jVhJazfAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a554a3fd8531e06d683f7b9e3782617bf4a3ae086c52643a69fbcd8c0b697933","last_reissued_at":"2026-07-05T07:50:55.847966Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:50:55.847966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2402.14270","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":"2026-07-05T07:50:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vmTbdVIOzv2gbL2zUAP7jeWPSjtFWP2Zayfjz7NLwiukbpjRRUklTenxHBWJMO8xSNni2sKbbiDy0SaQm0t2BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:56:06.963864Z"},"content_sha256":"e806cb5fbabcf7d8108f189c769f3338bf73b4dadcf446e3809713c4d6120f52","schema_version":"1.0","event_id":"sha256:e806cb5fbabcf7d8108f189c769f3338bf73b4dadcf446e3809713c4d6120f52"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:UVKKH7MFGHQG22B7POPDPATBPP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Take the Bull by the Horns: Hard Sample-Reweighted Continual Training Improves LLM Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daouda Sow, Junjie Yang, Mingyuan Zhou, Tianlong Chen, Xuxi Chen, Yingbin Liang, Zhangyang Wang, Zhendong Wang","submitted_at":"2024-02-22T04:10:57Z","abstract_excerpt":"In the rapidly advancing arena of large language models (LLMs), a key challenge is to enhance their capabilities amid a looming shortage of high-quality training data. Our study starts from an empirical strategy for the light continual training of LLMs using their original pre-training data sets, with a specific focus on selective retention of samples that incur moderately high losses. These samples are deemed informative and beneficial for model refinement, contrasting with the highest-loss samples, which would be discarded due to their correlation with data noise and complexity. We then form"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.14270","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/2402.14270/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":"2026-07-05T07:50:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o44fKaMHdcGqM1dZmn79S27mIze3SdnN91LJZpo6E8NEUehDmpxmY2j3MD+jT6E92jlATcN0u2NrH8FopwHVBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:56:06.964834Z"},"content_sha256":"c0e1a5ead4f7d2dc6504247479583d5439b4e0f80e16c995a4bc74065afab932","schema_version":"1.0","event_id":"sha256:c0e1a5ead4f7d2dc6504247479583d5439b4e0f80e16c995a4bc74065afab932"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UVKKH7MFGHQG22B7POPDPATBPP/bundle.json","state_url":"https://pith.science/pith/UVKKH7MFGHQG22B7POPDPATBPP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UVKKH7MFGHQG22B7POPDPATBPP/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":"2026-08-09T02:56:06Z","links":{"resolver":"https://pith.science/pith/UVKKH7MFGHQG22B7POPDPATBPP","bundle":"https://pith.science/pith/UVKKH7MFGHQG22B7POPDPATBPP/bundle.json","state":"https://pith.science/pith/UVKKH7MFGHQG22B7POPDPATBPP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UVKKH7MFGHQG22B7POPDPATBPP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:UVKKH7MFGHQG22B7POPDPATBPP","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":"f9b02e508659b8a20edf7a162d484037f4a51d6e18a872b75b4cc46bdd7a2b90","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-22T04:10:57Z","title_canon_sha256":"f95bb28957cc0cb08b299f7ea0c42c81321e759bf7442524fa1927f3688dc637"},"schema_version":"1.0","source":{"id":"2402.14270","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.14270","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"arxiv_version","alias_value":"2402.14270v2","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.14270","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_12","alias_value":"UVKKH7MFGHQG","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_16","alias_value":"UVKKH7MFGHQG22B7","created_at":"2026-07-05T07:50:55Z"},{"alias_kind":"pith_short_8","alias_value":"UVKKH7MF","created_at":"2026-07-05T07:50:55Z"}],"graph_snapshots":[{"event_id":"sha256:c0e1a5ead4f7d2dc6504247479583d5439b4e0f80e16c995a4bc74065afab932","target":"graph","created_at":"2026-07-05T07:50:55Z","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/2402.14270/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In the rapidly advancing arena of large language models (LLMs), a key challenge is to enhance their capabilities amid a looming shortage of high-quality training data. Our study starts from an empirical strategy for the light continual training of LLMs using their original pre-training data sets, with a specific focus on selective retention of samples that incur moderately high losses. These samples are deemed informative and beneficial for model refinement, contrasting with the highest-loss samples, which would be discarded due to their correlation with data noise and complexity. We then form","authors_text":"Daouda Sow, Junjie Yang, Mingyuan Zhou, Tianlong Chen, Xuxi Chen, Yingbin Liang, Zhangyang Wang, Zhendong Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-22T04:10:57Z","title":"Take the Bull by the Horns: Hard Sample-Reweighted Continual Training Improves LLM Generalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.14270","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:e806cb5fbabcf7d8108f189c769f3338bf73b4dadcf446e3809713c4d6120f52","target":"record","created_at":"2026-07-05T07:50:55Z","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":"f9b02e508659b8a20edf7a162d484037f4a51d6e18a872b75b4cc46bdd7a2b90","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-22T04:10:57Z","title_canon_sha256":"f95bb28957cc0cb08b299f7ea0c42c81321e759bf7442524fa1927f3688dc637"},"schema_version":"1.0","source":{"id":"2402.14270","kind":"arxiv","version":2}},"canonical_sha256":"a554a3fd8531e06d683f7b9e3782617bf4a3ae086c52643a69fbcd8c0b697933","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a554a3fd8531e06d683f7b9e3782617bf4a3ae086c52643a69fbcd8c0b697933","first_computed_at":"2026-07-05T07:50:55.847966Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:50:55.847966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6dRnpqOEJl/GnPySrH3Vs+0NuRt2piNfr6ytVs6aabak0Yk+D6VOJC/P+ejfufe8dWvxiOjCc4593jVhJazfAw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:50:55.848390Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.14270","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e806cb5fbabcf7d8108f189c769f3338bf73b4dadcf446e3809713c4d6120f52","sha256:c0e1a5ead4f7d2dc6504247479583d5439b4e0f80e16c995a4bc74065afab932"],"state_sha256":"f039c409ce73bf90772eb0df3a12bbc0b5d6e841c8d58ab05a25ffbe2beca45c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2OGTSp1Lkuys/nnjC7MKxUE9yNDz3A3mGDKqs0jxlW+QKeAOvKPJn6rWcySlGtARUO1sFs1yWVJLkjdiSCQHCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T02:56:06.971668Z","bundle_sha256":"c8e6a95ba58aeb86087c96c5bf9e61c78a1a64ea3ceacead51bd814f27ef2e85"}}