{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YQTEBTQA7AFIEIRVBW4H4NPHMM","short_pith_number":"pith:YQTEBTQA","schema_version":"1.0","canonical_sha256":"c42640ce00f80a8222350db87e35e7630fb63f4f54a28aa071c1ef93dd0da80e","source":{"kind":"arxiv","id":"2606.00160","version":1},"attestation_state":"computed","paper":{"title":"DataShield: Safety-degrading Data Filtering for LLM Benign Instruction Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CR","authors_text":"Jie Pan, Jinbiao Zhu, Junbo Zhang, Qianli Zhou, Wen Jiang, Xinyang Deng","submitted_at":"2026-05-29T09:04:27Z","abstract_excerpt":"Large language models (LLMs) suffer from degraded safety capabilities even when fine-tuned with benign datasets. However, existing methods for identifying safety-degrading samples in benign datasets suffer from high computational costs and significant noise issues. In this paper, we propose DataShield to efficiently and effectively identify potential safety-degrading samples. Our key intuition is based on the observation that benign fine-tuning increases the overall response compliance of LLMs. DataShield's key technical insight is to quantify each sample's contribution to the model's complian"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.00160","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2026-05-29T09:04:27Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"78cd9ec7c817493668dbf5d2459b7e83c92c6a933c18284145735c101066ac02","abstract_canon_sha256":"35970e1747bb7dfbc4dbce08f66697169f395c37de5e390e382f5e9bf429367f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:20.153702Z","signature_b64":"fqmY5Jh72WkfSzF43duFvynpyFX/YR5t+DmjAh+rg3mJAD03CJ95I2VjAUvVaXTZErsNHx4zMpCC6ZzgI874DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c42640ce00f80a8222350db87e35e7630fb63f4f54a28aa071c1ef93dd0da80e","last_reissued_at":"2026-06-02T01:03:20.153272Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:20.153272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DataShield: Safety-degrading Data Filtering for LLM Benign Instruction Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CR","authors_text":"Jie Pan, Jinbiao Zhu, Junbo Zhang, Qianli Zhou, Wen Jiang, Xinyang Deng","submitted_at":"2026-05-29T09:04:27Z","abstract_excerpt":"Large language models (LLMs) suffer from degraded safety capabilities even when fine-tuned with benign datasets. However, existing methods for identifying safety-degrading samples in benign datasets suffer from high computational costs and significant noise issues. In this paper, we propose DataShield to efficiently and effectively identify potential safety-degrading samples. Our key intuition is based on the observation that benign fine-tuning increases the overall response compliance of LLMs. DataShield's key technical insight is to quantify each sample's contribution to the model's complian"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00160","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/2606.00160/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.00160","created_at":"2026-06-02T01:03:20.153349+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.00160v1","created_at":"2026-06-02T01:03:20.153349+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00160","created_at":"2026-06-02T01:03:20.153349+00:00"},{"alias_kind":"pith_short_12","alias_value":"YQTEBTQA7AFI","created_at":"2026-06-02T01:03:20.153349+00:00"},{"alias_kind":"pith_short_16","alias_value":"YQTEBTQA7AFIEIRV","created_at":"2026-06-02T01:03:20.153349+00:00"},{"alias_kind":"pith_short_8","alias_value":"YQTEBTQA","created_at":"2026-06-02T01:03:20.153349+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM","json":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM.json","graph_json":"https://pith.science/api/pith-number/YQTEBTQA7AFIEIRVBW4H4NPHMM/graph.json","events_json":"https://pith.science/api/pith-number/YQTEBTQA7AFIEIRVBW4H4NPHMM/events.json","paper":"https://pith.science/paper/YQTEBTQA"},"agent_actions":{"view_html":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM","download_json":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM.json","view_paper":"https://pith.science/paper/YQTEBTQA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.00160&json=true","fetch_graph":"https://pith.science/api/pith-number/YQTEBTQA7AFIEIRVBW4H4NPHMM/graph.json","fetch_events":"https://pith.science/api/pith-number/YQTEBTQA7AFIEIRVBW4H4NPHMM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM/action/storage_attestation","attest_author":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM/action/author_attestation","sign_citation":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM/action/citation_signature","submit_replication":"https://pith.science/pith/YQTEBTQA7AFIEIRVBW4H4NPHMM/action/replication_record"}},"created_at":"2026-06-02T01:03:20.153349+00:00","updated_at":"2026-06-02T01:03:20.153349+00:00"}