{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:MLWZXDFUCKGV67IKBCX6JRKNKI","short_pith_number":"pith:MLWZXDFU","canonical_record":{"source":{"id":"2508.09101","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-12T17:29:20Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"1a2e74561b98df6b6702e0ce20634eab05a0ccf424ceb7c87b2a1a77f8d05d65","abstract_canon_sha256":"ebb86165df40f71899a4c23dfe4d4e17bc2f4a250e2fbc80da4936b9e8d67af4"},"schema_version":"1.0"},"canonical_sha256":"62ed9b8cb4128d5f7d0a08afe4c54d52282cf4a63074bcbd535a5c8a0fb8d78e","source":{"kind":"arxiv","id":"2508.09101","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.09101","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"arxiv_version","alias_value":"2508.09101v1","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.09101","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_12","alias_value":"MLWZXDFUCKGV","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_16","alias_value":"MLWZXDFUCKGV67IK","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_8","alias_value":"MLWZXDFU","created_at":"2026-07-05T11:52:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:MLWZXDFUCKGV67IKBCX6JRKNKI","target":"record","payload":{"canonical_record":{"source":{"id":"2508.09101","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-12T17:29:20Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"1a2e74561b98df6b6702e0ce20634eab05a0ccf424ceb7c87b2a1a77f8d05d65","abstract_canon_sha256":"ebb86165df40f71899a4c23dfe4d4e17bc2f4a250e2fbc80da4936b9e8d67af4"},"schema_version":"1.0"},"canonical_sha256":"62ed9b8cb4128d5f7d0a08afe4c54d52282cf4a63074bcbd535a5c8a0fb8d78e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:52:39.756265Z","signature_b64":"B8wa9lSXdYiuzTCg1cEQd6rWhu3465UJd6M/V/9vitWYrpsZHS6SeevE5w/wQx6wzgHmSh73pzhzu0vEB7bNCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"62ed9b8cb4128d5f7d0a08afe4c54d52282cf4a63074bcbd535a5c8a0fb8d78e","last_reissued_at":"2026-07-05T11:52:39.755764Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:52:39.755764Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.09101","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":"2026-07-05T11:52:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bihAHOiVzgS7xCC1KdcSnKuceG2wD5dNj9796PEaW+STyjMzTCcp0m5SvsLlj9xBMZtY9wINWlymPCR9MwswAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-11T04:17:50.580105Z"},"content_sha256":"3b94137d9c755cb98232a16247c7e842d549433dbf2cd7a8dd4105b1f4a4adf7","schema_version":"1.0","event_id":"sha256:3b94137d9c755cb98232a16247c7e842d549433dbf2cd7a8dd4105b1f4a4adf7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:MLWZXDFUCKGV67IKBCX6JRKNKI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AutoCodeBench: Large Language Models are Automatic Code Benchmark Generators","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":["cs.SE"],"primary_cat":"cs.CL","authors_text":"Ao Liu, Bohui Zhai, Chenchen Zhang, Fengzong Lian, Haotian Zhu, Hengyi Liu, Jason Chou, Jianwei Cai, Lingyun Tan, Speed Zhu, Tao Zhang, Wiggin Zhou, Yuchi Deng, Yue Mao, Zhiying Zeng, Ziyan Xu","submitted_at":"2025-08-12T17:29:20Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable capabilities across various domains, with code generation emerging as a key area of focus. While numerous benchmarks have been proposed to evaluate their code generation abilities, these benchmarks face several critical limitations. First, they often rely on manual annotations, which are time-consuming and difficult to scale across different programming languages and problem complexities. Second, most existing benchmarks focus primarily on Python, while the few multilingual benchmarks suffer from limited difficulty and uneven language d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.09101","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/2508.09101/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":"2026-07-05T11:52:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uZ/AaTlRDaMEtYhKgH64qE3JB70QHbWAw08xr90QSYVwjJFwq5nJH0S7GzGlNEGmiGPT0UPLuJ15BdwUzBIPBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-11T04:17:50.580505Z"},"content_sha256":"6b0aaced3df8736370182aef8f8cfdb8362eef073e63c68a6ba4df77e65806fb","schema_version":"1.0","event_id":"sha256:6b0aaced3df8736370182aef8f8cfdb8362eef073e63c68a6ba4df77e65806fb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/bundle.json","state_url":"https://pith.science/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/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":"2026-07-11T04:17:50Z","links":{"resolver":"https://pith.science/pith/MLWZXDFUCKGV67IKBCX6JRKNKI","bundle":"https://pith.science/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/bundle.json","state":"https://pith.science/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MLWZXDFUCKGV67IKBCX6JRKNKI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MLWZXDFUCKGV67IKBCX6JRKNKI","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":"ebb86165df40f71899a4c23dfe4d4e17bc2f4a250e2fbc80da4936b9e8d67af4","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-12T17:29:20Z","title_canon_sha256":"1a2e74561b98df6b6702e0ce20634eab05a0ccf424ceb7c87b2a1a77f8d05d65"},"schema_version":"1.0","source":{"id":"2508.09101","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.09101","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"arxiv_version","alias_value":"2508.09101v1","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.09101","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_12","alias_value":"MLWZXDFUCKGV","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_16","alias_value":"MLWZXDFUCKGV67IK","created_at":"2026-07-05T11:52:39Z"},{"alias_kind":"pith_short_8","alias_value":"MLWZXDFU","created_at":"2026-07-05T11:52:39Z"}],"graph_snapshots":[{"event_id":"sha256:6b0aaced3df8736370182aef8f8cfdb8362eef073e63c68a6ba4df77e65806fb","target":"graph","created_at":"2026-07-05T11:52:39Z","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/2508.09101/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable capabilities across various domains, with code generation emerging as a key area of focus. While numerous benchmarks have been proposed to evaluate their code generation abilities, these benchmarks face several critical limitations. First, they often rely on manual annotations, which are time-consuming and difficult to scale across different programming languages and problem complexities. Second, most existing benchmarks focus primarily on Python, while the few multilingual benchmarks suffer from limited difficulty and uneven language d","authors_text":"Ao Liu, Bohui Zhai, Chenchen Zhang, Fengzong Lian, Haotian Zhu, Hengyi Liu, Jason Chou, Jianwei Cai, Lingyun Tan, Speed Zhu, Tao Zhang, Wiggin Zhou, Yuchi Deng, Yue Mao, Zhiying Zeng, Ziyan Xu","cross_cats":["cs.SE"],"headline":"","license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-12T17:29:20Z","title":"AutoCodeBench: Large Language Models are Automatic Code Benchmark Generators"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.09101","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":"sha256:3b94137d9c755cb98232a16247c7e842d549433dbf2cd7a8dd4105b1f4a4adf7","target":"record","created_at":"2026-07-05T11:52:39Z","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":"ebb86165df40f71899a4c23dfe4d4e17bc2f4a250e2fbc80da4936b9e8d67af4","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-12T17:29:20Z","title_canon_sha256":"1a2e74561b98df6b6702e0ce20634eab05a0ccf424ceb7c87b2a1a77f8d05d65"},"schema_version":"1.0","source":{"id":"2508.09101","kind":"arxiv","version":1}},"canonical_sha256":"62ed9b8cb4128d5f7d0a08afe4c54d52282cf4a63074bcbd535a5c8a0fb8d78e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"62ed9b8cb4128d5f7d0a08afe4c54d52282cf4a63074bcbd535a5c8a0fb8d78e","first_computed_at":"2026-07-05T11:52:39.755764Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:52:39.755764Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"B8wa9lSXdYiuzTCg1cEQd6rWhu3465UJd6M/V/9vitWYrpsZHS6SeevE5w/wQx6wzgHmSh73pzhzu0vEB7bNCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:52:39.756265Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.09101","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3b94137d9c755cb98232a16247c7e842d549433dbf2cd7a8dd4105b1f4a4adf7","sha256:6b0aaced3df8736370182aef8f8cfdb8362eef073e63c68a6ba4df77e65806fb"],"state_sha256":"cccd52443283b24336288749f6a4113f0b829bd20f4ac922ffd637028a8ade57"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KqPhYcuTd11y7KiW9suL9ys126NaZPBwx68M/TwmL32rpHF8SbMTZfNAxpCLZvA4Iy9Jo3q9E73VlzBVkvfyCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-11T04:17:50.582637Z","bundle_sha256":"710d58c35c1bdb46e11b1c4e6abcca2861fb0f64852df626e808b30a78498821"}}