{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:ESHC47X7ZVKYU34AVSZHULZZ34","short_pith_number":"pith:ESHC47X7","canonical_record":{"source":{"id":"2406.12952","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-18T14:54:37Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"e38cc0ccf8c17bbf24bdfd4ac3a98db804ce3a61fd49cbf1e6350eb501a7e3f6","abstract_canon_sha256":"bf3e860af563d9eb319cdba459900fedc03f8abcbeea72173105079309622685"},"schema_version":"1.0"},"canonical_sha256":"248e2e7effcd558a6f80acb27a2f39df3f55e0a6a3a64c1d4010427675f3d765","source":{"kind":"arxiv","id":"2406.12952","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12952","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12952v3","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12952","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_12","alias_value":"ESHC47X7ZVKY","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_16","alias_value":"ESHC47X7ZVKYU34A","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_8","alias_value":"ESHC47X7","created_at":"2026-07-05T10:10:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:ESHC47X7ZVKYU34AVSZHULZZ34","target":"record","payload":{"canonical_record":{"source":{"id":"2406.12952","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-18T14:54:37Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"e38cc0ccf8c17bbf24bdfd4ac3a98db804ce3a61fd49cbf1e6350eb501a7e3f6","abstract_canon_sha256":"bf3e860af563d9eb319cdba459900fedc03f8abcbeea72173105079309622685"},"schema_version":"1.0"},"canonical_sha256":"248e2e7effcd558a6f80acb27a2f39df3f55e0a6a3a64c1d4010427675f3d765","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:41.287678Z","signature_b64":"egmoprRZOc1pCXCLMLCE5okk/eCdIKsTXimmdHTYgpGU+Vq0U6r9sxbyM2SLLqG7lAsSKlkiJsVGLuhpMWouCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"248e2e7effcd558a6f80acb27a2f39df3f55e0a6a3a64c1d4010427675f3d765","last_reissued_at":"2026-07-05T10:10:41.287211Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:41.287211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.12952","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:10:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BQdpnIgZicVHH3TfjQAFxBBlsC15rjDGmCYETvhJxxFEYiVkCfbKWco2hX6afVajS3Om7R8QmxuJrbka9ps1DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:37:46.664494Z"},"content_sha256":"4b6e6f6af1b5a87f07ebe2f41b55ebb69e580106cdb43ef28fa5b7c0bc991aff","schema_version":"1.0","event_id":"sha256:4b6e6f6af1b5a87f07ebe2f41b55ebb69e580106cdb43ef28fa5b7c0bc991aff"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:ESHC47X7ZVKYU34AVSZHULZZ34","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SWT-Bench: Testing and Validating Real-World Bug-Fixes with Code Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.SE","authors_text":"Jingxuan He, Mark Niklas M\\\"uller, Martin Vechev, Niels M\\\"undler","submitted_at":"2024-06-18T14:54:37Z","abstract_excerpt":"Rigorous software testing is crucial for developing and maintaining high-quality code, making automated test generation a promising avenue for both improving software quality and boosting the effectiveness of code generation methods. However, while code generation with Large Language Models (LLMs) is an extraordinarily active research area, test generation remains relatively unexplored. We address this gap and investigate the capability of LLM-based Code Agents to formalize user issues into test cases. To this end, we propose a novel benchmark based on popular GitHub repositories, containing r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12952","kind":"arxiv","version":3},"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/2406.12952/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":"2026-07-05T10:10:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ILZxelcY7eTKf5nzc2CbXBaWzOg62NzHdmtQfgw5FwZg0XTupJ3Pyp+KfRUsYzuxhazHgTQLZKL2nAlUS6ALCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:37:46.664869Z"},"content_sha256":"0dd3cc39c9e6a35052509d44685acb1a487e4b13caca25d35b2f59388fc2124b","schema_version":"1.0","event_id":"sha256:0dd3cc39c9e6a35052509d44685acb1a487e4b13caca25d35b2f59388fc2124b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ESHC47X7ZVKYU34AVSZHULZZ34/bundle.json","state_url":"https://pith.science/pith/ESHC47X7ZVKYU34AVSZHULZZ34/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ESHC47X7ZVKYU34AVSZHULZZ34/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":"2026-07-07T11:37:46Z","links":{"resolver":"https://pith.science/pith/ESHC47X7ZVKYU34AVSZHULZZ34","bundle":"https://pith.science/pith/ESHC47X7ZVKYU34AVSZHULZZ34/bundle.json","state":"https://pith.science/pith/ESHC47X7ZVKYU34AVSZHULZZ34/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ESHC47X7ZVKYU34AVSZHULZZ34/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:ESHC47X7ZVKYU34AVSZHULZZ34","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":"bf3e860af563d9eb319cdba459900fedc03f8abcbeea72173105079309622685","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-18T14:54:37Z","title_canon_sha256":"e38cc0ccf8c17bbf24bdfd4ac3a98db804ce3a61fd49cbf1e6350eb501a7e3f6"},"schema_version":"1.0","source":{"id":"2406.12952","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12952","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12952v3","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12952","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_12","alias_value":"ESHC47X7ZVKY","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_16","alias_value":"ESHC47X7ZVKYU34A","created_at":"2026-07-05T10:10:41Z"},{"alias_kind":"pith_short_8","alias_value":"ESHC47X7","created_at":"2026-07-05T10:10:41Z"}],"graph_snapshots":[{"event_id":"sha256:0dd3cc39c9e6a35052509d44685acb1a487e4b13caca25d35b2f59388fc2124b","target":"graph","created_at":"2026-07-05T10:10:41Z","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/2406.12952/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Rigorous software testing is crucial for developing and maintaining high-quality code, making automated test generation a promising avenue for both improving software quality and boosting the effectiveness of code generation methods. However, while code generation with Large Language Models (LLMs) is an extraordinarily active research area, test generation remains relatively unexplored. We address this gap and investigate the capability of LLM-based Code Agents to formalize user issues into test cases. To this end, we propose a novel benchmark based on popular GitHub repositories, containing r","authors_text":"Jingxuan He, Mark Niklas M\\\"uller, Martin Vechev, Niels M\\\"undler","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-18T14:54:37Z","title":"SWT-Bench: Testing and Validating Real-World Bug-Fixes with Code Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12952","kind":"arxiv","version":3},"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":"sha256:4b6e6f6af1b5a87f07ebe2f41b55ebb69e580106cdb43ef28fa5b7c0bc991aff","target":"record","created_at":"2026-07-05T10:10:41Z","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":"bf3e860af563d9eb319cdba459900fedc03f8abcbeea72173105079309622685","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-06-18T14:54:37Z","title_canon_sha256":"e38cc0ccf8c17bbf24bdfd4ac3a98db804ce3a61fd49cbf1e6350eb501a7e3f6"},"schema_version":"1.0","source":{"id":"2406.12952","kind":"arxiv","version":3}},"canonical_sha256":"248e2e7effcd558a6f80acb27a2f39df3f55e0a6a3a64c1d4010427675f3d765","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"248e2e7effcd558a6f80acb27a2f39df3f55e0a6a3a64c1d4010427675f3d765","first_computed_at":"2026-07-05T10:10:41.287211Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:10:41.287211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"egmoprRZOc1pCXCLMLCE5okk/eCdIKsTXimmdHTYgpGU+Vq0U6r9sxbyM2SLLqG7lAsSKlkiJsVGLuhpMWouCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:10:41.287678Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.12952","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4b6e6f6af1b5a87f07ebe2f41b55ebb69e580106cdb43ef28fa5b7c0bc991aff","sha256:0dd3cc39c9e6a35052509d44685acb1a487e4b13caca25d35b2f59388fc2124b"],"state_sha256":"b95fd6ce60a64f3e79cde6358bef257b0968e3c15e2e49c78cc69af1292da60c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cJ8aG2CDs1t0jjNcJ7vysGgDb50Uo8lhwuvR+VGaEoImg778BHlVYEsD1KkNeKcxmEHauFBMuXIScGFxs3s0Bw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T11:37:46.666787Z","bundle_sha256":"d3c9078f3000069890ba884e2ee2a4310bd2129675c8e21a2ad45925c900b6f5"}}