{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:654EYS7ERDOJM7X3WH2JYA6XM7","short_pith_number":"pith:654EYS7E","canonical_record":{"source":{"id":"2505.12331","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-18T09:53:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"95da90041d6b3d25e17d63492368c51407b08878abc7056bfaf215427764b08e","abstract_canon_sha256":"7967828f73286bf6c9be1559d3e42de5c78787185e4392e12f66db524154b904"},"schema_version":"1.0"},"canonical_sha256":"f7784c4be488dc967efbb1f49c03d767d4e9cffac8ca5b3cbe29077138b97d79","source":{"kind":"arxiv","id":"2505.12331","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12331","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12331v2","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12331","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_12","alias_value":"654EYS7ERDOJ","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_16","alias_value":"654EYS7ERDOJM7X3","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_8","alias_value":"654EYS7E","created_at":"2026-07-05T11:05:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:654EYS7ERDOJM7X3WH2JYA6XM7","target":"record","payload":{"canonical_record":{"source":{"id":"2505.12331","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-18T09:53:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"95da90041d6b3d25e17d63492368c51407b08878abc7056bfaf215427764b08e","abstract_canon_sha256":"7967828f73286bf6c9be1559d3e42de5c78787185e4392e12f66db524154b904"},"schema_version":"1.0"},"canonical_sha256":"f7784c4be488dc967efbb1f49c03d767d4e9cffac8ca5b3cbe29077138b97d79","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:45.281002Z","signature_b64":"hAX/IVYiOcDz3xt7hi3LvkfUEqP8bIAc0UId3MEiIyiI0dIWgmh1f+Lxg3UKKHkCuVskjVA9QrFJST+149aQBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f7784c4be488dc967efbb1f49c03d767d4e9cffac8ca5b3cbe29077138b97d79","last_reissued_at":"2026-07-05T11:05:45.280531Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:45.280531Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.12331","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":"2026-07-05T11:05:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bjDslPsG34ZAb0r0rItO3ue37Qh5eSyqp6Eu8wa9BaZu6V+IyBZba6xtTtMAMl9Z6MgV8GdeX4wMQcM6deZ2BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T02:11:27.173881Z"},"content_sha256":"df81b6d31914b5fdd418404eb4940d737d3923faf08aba6fd75ea558b58c3d24","schema_version":"1.0","event_id":"sha256:df81b6d31914b5fdd418404eb4940d737d3923faf08aba6fd75ea558b58c3d24"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:654EYS7ERDOJM7X3WH2JYA6XM7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"OSS-Bench: Benchmark Generator for Coding LLMs","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.SE","authors_text":"Roland Yap, Yuancheng Jiang, Zhenkai Liang","submitted_at":"2025-05-18T09:53:51Z","abstract_excerpt":"In light of the rapid adoption of AI coding assistants, LLM-assisted development has become increasingly prevalent, creating an urgent need for robust evaluation of generated code quality. Existing benchmarks often require extensive manual effort to create static datasets, rely on indirect or insufficiently challenging tasks, depend on non-scalable ground truth, or neglect critical low-level security evaluations, particularly memory-safety issues. In this work, we introduce OSS-Bench, a benchmark generator that automatically constructs large-scale, live evaluation tasks from real-world open-so"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12331","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/2505.12331/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":"2026-07-05T11:05:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WZ9rNWDp6AJIEQnJpGASSjD2grvbuKZx3Ev4VY9nUj/FnrWyoVOtzJGzB4S7EEuX+O2EEEjCmMYwUli8JMsFAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T02:11:27.174635Z"},"content_sha256":"5bb56ea2e8d0c65c2c16ae8b92c71f7056f07c523e9bedb238672f82ab1f1530","schema_version":"1.0","event_id":"sha256:5bb56ea2e8d0c65c2c16ae8b92c71f7056f07c523e9bedb238672f82ab1f1530"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/654EYS7ERDOJM7X3WH2JYA6XM7/bundle.json","state_url":"https://pith.science/pith/654EYS7ERDOJM7X3WH2JYA6XM7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/654EYS7ERDOJM7X3WH2JYA6XM7/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":"2026-08-19T02:11:27Z","links":{"resolver":"https://pith.science/pith/654EYS7ERDOJM7X3WH2JYA6XM7","bundle":"https://pith.science/pith/654EYS7ERDOJM7X3WH2JYA6XM7/bundle.json","state":"https://pith.science/pith/654EYS7ERDOJM7X3WH2JYA6XM7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/654EYS7ERDOJM7X3WH2JYA6XM7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:654EYS7ERDOJM7X3WH2JYA6XM7","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":"7967828f73286bf6c9be1559d3e42de5c78787185e4392e12f66db524154b904","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-18T09:53:51Z","title_canon_sha256":"95da90041d6b3d25e17d63492368c51407b08878abc7056bfaf215427764b08e"},"schema_version":"1.0","source":{"id":"2505.12331","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12331","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12331v2","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12331","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_12","alias_value":"654EYS7ERDOJ","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_16","alias_value":"654EYS7ERDOJM7X3","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_8","alias_value":"654EYS7E","created_at":"2026-07-05T11:05:45Z"}],"graph_snapshots":[{"event_id":"sha256:5bb56ea2e8d0c65c2c16ae8b92c71f7056f07c523e9bedb238672f82ab1f1530","target":"graph","created_at":"2026-07-05T11:05:45Z","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/2505.12331/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In light of the rapid adoption of AI coding assistants, LLM-assisted development has become increasingly prevalent, creating an urgent need for robust evaluation of generated code quality. Existing benchmarks often require extensive manual effort to create static datasets, rely on indirect or insufficiently challenging tasks, depend on non-scalable ground truth, or neglect critical low-level security evaluations, particularly memory-safety issues. In this work, we introduce OSS-Bench, a benchmark generator that automatically constructs large-scale, live evaluation tasks from real-world open-so","authors_text":"Roland Yap, Yuancheng Jiang, Zhenkai Liang","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-18T09:53:51Z","title":"OSS-Bench: Benchmark Generator for Coding LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12331","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:df81b6d31914b5fdd418404eb4940d737d3923faf08aba6fd75ea558b58c3d24","target":"record","created_at":"2026-07-05T11:05:45Z","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":"7967828f73286bf6c9be1559d3e42de5c78787185e4392e12f66db524154b904","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.SE","submitted_at":"2025-05-18T09:53:51Z","title_canon_sha256":"95da90041d6b3d25e17d63492368c51407b08878abc7056bfaf215427764b08e"},"schema_version":"1.0","source":{"id":"2505.12331","kind":"arxiv","version":2}},"canonical_sha256":"f7784c4be488dc967efbb1f49c03d767d4e9cffac8ca5b3cbe29077138b97d79","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f7784c4be488dc967efbb1f49c03d767d4e9cffac8ca5b3cbe29077138b97d79","first_computed_at":"2026-07-05T11:05:45.280531Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:45.280531Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hAX/IVYiOcDz3xt7hi3LvkfUEqP8bIAc0UId3MEiIyiI0dIWgmh1f+Lxg3UKKHkCuVskjVA9QrFJST+149aQBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:45.281002Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12331","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:df81b6d31914b5fdd418404eb4940d737d3923faf08aba6fd75ea558b58c3d24","sha256:5bb56ea2e8d0c65c2c16ae8b92c71f7056f07c523e9bedb238672f82ab1f1530"],"state_sha256":"08d89d03882433b565a3142149865d93d70314d7312070c8a483236106752eef"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jHMpwXelvi/mS6drZLef3gRVSwAWUhbwQwhzvWThew90AyTAJfMxoygDplYRpKNaMMIx12qVR3rsN2x+K9eBBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T02:11:27.179812Z","bundle_sha256":"1f856faa39c59d1879bf6d81c6142280cc09eba97c8fb5702ef4a30db083a865"}}