{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:VCAQDLJLXSHOEUV6W3REHR7NUL","short_pith_number":"pith:VCAQDLJL","canonical_record":{"source":{"id":"2410.12999","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T19:56:22Z","cross_cats_sorted":[],"title_canon_sha256":"f01093045ed46cb4911964da2d337cce3a3daa280077ecce4f2b9d67db1d274a","abstract_canon_sha256":"dc5d67ae08521857b0e6cea332f4e7b0c77079af3e93d840c73c2ad0adeaf81c"},"schema_version":"1.0"},"canonical_sha256":"a88101ad2bbc8ee252beb6e243c7eda2f0c06f2e06def21654b413d04c900b62","source":{"kind":"arxiv","id":"2410.12999","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.12999","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"arxiv_version","alias_value":"2410.12999v2","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.12999","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_12","alias_value":"VCAQDLJLXSHO","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_16","alias_value":"VCAQDLJLXSHOEUV6","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_8","alias_value":"VCAQDLJL","created_at":"2026-07-05T11:21:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:VCAQDLJLXSHOEUV6W3REHR7NUL","target":"record","payload":{"canonical_record":{"source":{"id":"2410.12999","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T19:56:22Z","cross_cats_sorted":[],"title_canon_sha256":"f01093045ed46cb4911964da2d337cce3a3daa280077ecce4f2b9d67db1d274a","abstract_canon_sha256":"dc5d67ae08521857b0e6cea332f4e7b0c77079af3e93d840c73c2ad0adeaf81c"},"schema_version":"1.0"},"canonical_sha256":"a88101ad2bbc8ee252beb6e243c7eda2f0c06f2e06def21654b413d04c900b62","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:21:37.150997Z","signature_b64":"ZHByr3F9DroLBsPkeFV9kehW491Ns9sQZ+i/7A9vte+rqAG5qUM6vjy84YM5kqcbN0vxegWchrlfM9Y4RT2iBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a88101ad2bbc8ee252beb6e243c7eda2f0c06f2e06def21654b413d04c900b62","last_reissued_at":"2026-07-05T11:21:37.150492Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:21:37.150492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.12999","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":"2026-07-05T11:21:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P1uhEK9GaS0hMCNhCkwT+RMJ5WK81QGZB54rYTx91rRmqFulo+8is1RoKOyrnzNKsWkLyB3qalqxuTSigc/5Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T09:30:40.256360Z"},"content_sha256":"77486031c10035b22af22310f31c660eda4581ff917dd2a14076032c9e8b1c2a","schema_version":"1.0","event_id":"sha256:77486031c10035b22af22310f31c660eda4581ff917dd2a14076032c9e8b1c2a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:VCAQDLJLXSHOEUV6W3REHR7NUL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"POROver: Improving Safety and Reducing Overrefusal in Large Language Models with Overgeneration and Preference Optimization","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alon Benhaim, Batuhan K. Karaman, Ishmam Zabir, Mert R. Sabuncu, Vishrav Chaudhary, Xia Song","submitted_at":"2024-10-16T19:56:22Z","abstract_excerpt":"Achieving both high safety and high usefulness simultaneously in large language models has become a critical challenge in recent years.Models often exhibit unsafe behavior or adopt an overly cautious approach leading to frequent overrefusal of benign prompts, which reduces their usefulness. A major factor underlying these behaviors is how the models are finetuned and aligned, particularly the nature and extent of the data used.In this work, we examine how overgenerating finetuning data with advanced teacher models (e.g., GPT-4o)-covering both general-purpose and toxic prompts-affects safety an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.12999","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/2410.12999/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":"2026-07-05T11:21:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CjnKkHucg1phsDb1TxNuEzUr8d1Lut0IAPyeIEmUeCWuRuQlpyBIa6vptTmy3YAmnLdQ5OPfTbxuuCGiAmCNDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T09:30:40.257357Z"},"content_sha256":"8ab6c8f0be5d89849ba7c70ec56724b9befa12c0cbade4966060ebe948591d95","schema_version":"1.0","event_id":"sha256:8ab6c8f0be5d89849ba7c70ec56724b9befa12c0cbade4966060ebe948591d95"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/bundle.json","state_url":"https://pith.science/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/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":"2026-08-17T09:30:40Z","links":{"resolver":"https://pith.science/pith/VCAQDLJLXSHOEUV6W3REHR7NUL","bundle":"https://pith.science/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/bundle.json","state":"https://pith.science/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VCAQDLJLXSHOEUV6W3REHR7NUL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:VCAQDLJLXSHOEUV6W3REHR7NUL","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":"dc5d67ae08521857b0e6cea332f4e7b0c77079af3e93d840c73c2ad0adeaf81c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T19:56:22Z","title_canon_sha256":"f01093045ed46cb4911964da2d337cce3a3daa280077ecce4f2b9d67db1d274a"},"schema_version":"1.0","source":{"id":"2410.12999","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.12999","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"arxiv_version","alias_value":"2410.12999v2","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.12999","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_12","alias_value":"VCAQDLJLXSHO","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_16","alias_value":"VCAQDLJLXSHOEUV6","created_at":"2026-07-05T11:21:37Z"},{"alias_kind":"pith_short_8","alias_value":"VCAQDLJL","created_at":"2026-07-05T11:21:37Z"}],"graph_snapshots":[{"event_id":"sha256:8ab6c8f0be5d89849ba7c70ec56724b9befa12c0cbade4966060ebe948591d95","target":"graph","created_at":"2026-07-05T11:21:37Z","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/2410.12999/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Achieving both high safety and high usefulness simultaneously in large language models has become a critical challenge in recent years.Models often exhibit unsafe behavior or adopt an overly cautious approach leading to frequent overrefusal of benign prompts, which reduces their usefulness. A major factor underlying these behaviors is how the models are finetuned and aligned, particularly the nature and extent of the data used.In this work, we examine how overgenerating finetuning data with advanced teacher models (e.g., GPT-4o)-covering both general-purpose and toxic prompts-affects safety an","authors_text":"Alon Benhaim, Batuhan K. Karaman, Ishmam Zabir, Mert R. Sabuncu, Vishrav Chaudhary, Xia Song","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T19:56:22Z","title":"POROver: Improving Safety and Reducing Overrefusal in Large Language Models with Overgeneration and Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.12999","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":"sha256:77486031c10035b22af22310f31c660eda4581ff917dd2a14076032c9e8b1c2a","target":"record","created_at":"2026-07-05T11:21:37Z","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":"dc5d67ae08521857b0e6cea332f4e7b0c77079af3e93d840c73c2ad0adeaf81c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T19:56:22Z","title_canon_sha256":"f01093045ed46cb4911964da2d337cce3a3daa280077ecce4f2b9d67db1d274a"},"schema_version":"1.0","source":{"id":"2410.12999","kind":"arxiv","version":2}},"canonical_sha256":"a88101ad2bbc8ee252beb6e243c7eda2f0c06f2e06def21654b413d04c900b62","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a88101ad2bbc8ee252beb6e243c7eda2f0c06f2e06def21654b413d04c900b62","first_computed_at":"2026-07-05T11:21:37.150492Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:21:37.150492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZHByr3F9DroLBsPkeFV9kehW491Ns9sQZ+i/7A9vte+rqAG5qUM6vjy84YM5kqcbN0vxegWchrlfM9Y4RT2iBg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:21:37.150997Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.12999","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:77486031c10035b22af22310f31c660eda4581ff917dd2a14076032c9e8b1c2a","sha256:8ab6c8f0be5d89849ba7c70ec56724b9befa12c0cbade4966060ebe948591d95"],"state_sha256":"0e5a604614e3abd56b3874dff0dd052d258bf52fb068c0962ee5dd342d338186"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"85QTIxQGYIXqvy1rWVVJK0jr5UbqrHxWgUHO+jF90ZXmUx8caHAVUb8+HcaismSZQZeZig1pWXQtuGt7ZuluCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T09:30:40.264500Z","bundle_sha256":"f108984faceacbfe8f7774d3e03ba13f60a4e6bccd61a5c2d8679bb1598efed4"}}