{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:SWUEFP6FVA6X3BYKH5AWJ45T4V","short_pith_number":"pith:SWUEFP6F","canonical_record":{"source":{"id":"2505.10494","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-15T16:53:41Z","cross_cats_sorted":[],"title_canon_sha256":"94665fb83d99afd4d42a0e16e9be3f5ea65b7851797ce335472ba158f0c69acc","abstract_canon_sha256":"c7ead6efe304f230f8a4a4b4fc69f94c271ec4999b85b134a14c55260ed5e881"},"schema_version":"1.0"},"canonical_sha256":"95a842bfc5a83d7d870a3f4164f3b3e55cc8f11161f1e36c0d28abd73a6bfd62","source":{"kind":"arxiv","id":"2505.10494","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.10494","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"arxiv_version","alias_value":"2505.10494v1","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.10494","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_12","alias_value":"SWUEFP6FVA6X","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_16","alias_value":"SWUEFP6FVA6X3BYK","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_8","alias_value":"SWUEFP6F","created_at":"2026-07-05T11:03:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:SWUEFP6FVA6X3BYKH5AWJ45T4V","target":"record","payload":{"canonical_record":{"source":{"id":"2505.10494","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-15T16:53:41Z","cross_cats_sorted":[],"title_canon_sha256":"94665fb83d99afd4d42a0e16e9be3f5ea65b7851797ce335472ba158f0c69acc","abstract_canon_sha256":"c7ead6efe304f230f8a4a4b4fc69f94c271ec4999b85b134a14c55260ed5e881"},"schema_version":"1.0"},"canonical_sha256":"95a842bfc5a83d7d870a3f4164f3b3e55cc8f11161f1e36c0d28abd73a6bfd62","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:03:39.101085Z","signature_b64":"+k7qOOSpiviyRo+23UaT0paaFZHaS4yQZigkT//LiLVBR3yMNBcZeB2bY14qnLl9RoOFpSRtvqhX8RLBOuUUBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"95a842bfc5a83d7d870a3f4164f3b3e55cc8f11161f1e36c0d28abd73a6bfd62","last_reissued_at":"2026-07-05T11:03:39.100637Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:03:39.100637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.10494","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:03:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3voiCsQtaRu8HKldEWbtXFNco6V5e3m2ee2nwYkdZbm6m1dUjBFmhun++FjfjJxBmJbm2VHTrSqWuuU7H+3VDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T11:45:09.455100Z"},"content_sha256":"587c352926efc155a2df5951070e9c7951dc92c0f426aad2e8f1f5ab9df3f569","schema_version":"1.0","event_id":"sha256:587c352926efc155a2df5951070e9c7951dc92c0f426aad2e8f1f5ab9df3f569"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:SWUEFP6FVA6X3BYKH5AWJ45T4V","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can You Really Trust Code Copilots? Evaluating Large Language Models from a Code Security Perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Shikun Zhang, Wei Ye, Xiao Deng, Yutao Mou, Yuxiao Luo","submitted_at":"2025-05-15T16:53:41Z","abstract_excerpt":"Code security and usability are both essential for various coding assistant applications driven by large language models (LLMs). Current code security benchmarks focus solely on single evaluation task and paradigm, such as code completion and generation, lacking comprehensive assessment across dimensions like secure code generation, vulnerability repair and discrimination. In this paper, we first propose CoV-Eval, a multi-task benchmark covering various tasks such as code completion, vulnerability repair, vulnerability detection and classification, for comprehensive evaluation of LLM code secu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.10494","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/2505.10494/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":"2026-07-05T11:03:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pIecJsO12j3AmdPLi8KKcJUH7BHg2XMKrCLvrFH/AtUSJdUBWtXnwb7prSJL7BeMuU6+0XoKi3wgvYdNjGgWDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T11:45:09.455483Z"},"content_sha256":"e07f2944f7ed45e39b9d68223dbfa1c95da0fd5f29e3e27363ebad7478e64859","schema_version":"1.0","event_id":"sha256:e07f2944f7ed45e39b9d68223dbfa1c95da0fd5f29e3e27363ebad7478e64859"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/bundle.json","state_url":"https://pith.science/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/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":"2026-07-06T11:45:09Z","links":{"resolver":"https://pith.science/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V","bundle":"https://pith.science/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/bundle.json","state":"https://pith.science/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SWUEFP6FVA6X3BYKH5AWJ45T4V/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:SWUEFP6FVA6X3BYKH5AWJ45T4V","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":"c7ead6efe304f230f8a4a4b4fc69f94c271ec4999b85b134a14c55260ed5e881","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-15T16:53:41Z","title_canon_sha256":"94665fb83d99afd4d42a0e16e9be3f5ea65b7851797ce335472ba158f0c69acc"},"schema_version":"1.0","source":{"id":"2505.10494","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.10494","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"arxiv_version","alias_value":"2505.10494v1","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.10494","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_12","alias_value":"SWUEFP6FVA6X","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_16","alias_value":"SWUEFP6FVA6X3BYK","created_at":"2026-07-05T11:03:39Z"},{"alias_kind":"pith_short_8","alias_value":"SWUEFP6F","created_at":"2026-07-05T11:03:39Z"}],"graph_snapshots":[{"event_id":"sha256:e07f2944f7ed45e39b9d68223dbfa1c95da0fd5f29e3e27363ebad7478e64859","target":"graph","created_at":"2026-07-05T11:03:39Z","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/2505.10494/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Code security and usability are both essential for various coding assistant applications driven by large language models (LLMs). Current code security benchmarks focus solely on single evaluation task and paradigm, such as code completion and generation, lacking comprehensive assessment across dimensions like secure code generation, vulnerability repair and discrimination. In this paper, we first propose CoV-Eval, a multi-task benchmark covering various tasks such as code completion, vulnerability repair, vulnerability detection and classification, for comprehensive evaluation of LLM code secu","authors_text":"Shikun Zhang, Wei Ye, Xiao Deng, Yutao Mou, Yuxiao Luo","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-15T16:53:41Z","title":"Can You Really Trust Code Copilots? Evaluating Large Language Models from a Code Security Perspective"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.10494","kind":"arxiv","version":1},"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":"sha256:587c352926efc155a2df5951070e9c7951dc92c0f426aad2e8f1f5ab9df3f569","target":"record","created_at":"2026-07-05T11:03:39Z","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":"c7ead6efe304f230f8a4a4b4fc69f94c271ec4999b85b134a14c55260ed5e881","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-15T16:53:41Z","title_canon_sha256":"94665fb83d99afd4d42a0e16e9be3f5ea65b7851797ce335472ba158f0c69acc"},"schema_version":"1.0","source":{"id":"2505.10494","kind":"arxiv","version":1}},"canonical_sha256":"95a842bfc5a83d7d870a3f4164f3b3e55cc8f11161f1e36c0d28abd73a6bfd62","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"95a842bfc5a83d7d870a3f4164f3b3e55cc8f11161f1e36c0d28abd73a6bfd62","first_computed_at":"2026-07-05T11:03:39.100637Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:03:39.100637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+k7qOOSpiviyRo+23UaT0paaFZHaS4yQZigkT//LiLVBR3yMNBcZeB2bY14qnLl9RoOFpSRtvqhX8RLBOuUUBg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:03:39.101085Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.10494","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:587c352926efc155a2df5951070e9c7951dc92c0f426aad2e8f1f5ab9df3f569","sha256:e07f2944f7ed45e39b9d68223dbfa1c95da0fd5f29e3e27363ebad7478e64859"],"state_sha256":"2eab8ac36722b6124e389a0f056c0d2e2e13f29b97569385fd21d56e54a6759f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VubhpR1vu3vr0TAx08LujRx4BulOGaikELlleYhbx8tpNcZ+qJ0Gf/mSVt8sDP2koNT3ZTrJKTx7vVc9TiyGAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T11:45:09.457453Z","bundle_sha256":"0d0656c3092025adfc2d3f7546cb1c089a2729267205a556847182ba6b3b289f"}}