{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:N75JPUYRSEA6E72RKYYPY4SPTK","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":"eeeeb21ebc694961cf736d79f8eef058f8647377f624c71e6107c8ebfd6b2f12","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-09-01T14:13:34Z","title_canon_sha256":"f2db0372e4b3b178e847b818d6d64dd899fc6287b16ab2821a5857a5001a162b"},"schema_version":"1.0","source":{"id":"2509.01494","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.01494","created_at":"2026-06-08T01:03:47Z"},{"alias_kind":"arxiv_version","alias_value":"2509.01494v2","created_at":"2026-06-08T01:03:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.01494","created_at":"2026-06-08T01:03:47Z"},{"alias_kind":"pith_short_12","alias_value":"N75JPUYRSEA6","created_at":"2026-06-08T01:03:47Z"},{"alias_kind":"pith_short_16","alias_value":"N75JPUYRSEA6E72R","created_at":"2026-06-08T01:03:47Z"},{"alias_kind":"pith_short_8","alias_value":"N75JPUYR","created_at":"2026-06-08T01:03:47Z"}],"graph_snapshots":[{"event_id":"sha256:6420ba14ded845c8ff2647322a296d7409beb2f48afb88af9a9f03ca7ff47eb1","target":"graph","created_at":"2026-06-08T01:03:47Z","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/2509.01494/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Automated Code Review (ACR) is crucial for software quality, yet existing benchmarks often fail to reflect real-world complexities, hindering the evaluation of modern Large Language Models (LLMs). Current benchmarks frequently focus on fine-grained code units, lack complete project context, and use inadequate evaluation metrics. To address these limitations, we introduce SWRBench , a new benchmark comprising 1000 manually verified Pull Requests (PRs) from GitHub, offering PR-centric review with full project context. SWRBench employs an objective LLM-based evaluation method that aligns strongly","authors_text":"Kaicheng Sun, Keke Han, Ruikai Shi, Rui Xie, Shikun Zhang, Wei Ye, Yidong Wang, Yixin Li, Zhengran Zeng, Zhuohao Yu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-09-01T14:13:34Z","title":"SWR-Bench: Assessing LLM Performance in Real-World Code Review Comment Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.01494","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:ef9a5b1edd231d21981de6014ce26e9d397a4cf791d777951a6d9aff455198f6","target":"record","created_at":"2026-06-08T01:03:47Z","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":"eeeeb21ebc694961cf736d79f8eef058f8647377f624c71e6107c8ebfd6b2f12","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-09-01T14:13:34Z","title_canon_sha256":"f2db0372e4b3b178e847b818d6d64dd899fc6287b16ab2821a5857a5001a162b"},"schema_version":"1.0","source":{"id":"2509.01494","kind":"arxiv","version":2}},"canonical_sha256":"6ffa97d3119101e27f515630fc724f9a9e33d8c9aa6cdc2531e1017ec6b73828","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6ffa97d3119101e27f515630fc724f9a9e33d8c9aa6cdc2531e1017ec6b73828","first_computed_at":"2026-06-08T01:03:47.947761Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-08T01:03:47.947761Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IQ6fPZ4pF8RLuwqWsLJKF2k2s47nuyoF3wuJ0bpzcT7mwvFDOwCCowGuVmf0UBT9+ptkdIbEE8lG/9gEMh3yAg==","signature_status":"signed_v1","signed_at":"2026-06-08T01:03:47.948580Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.01494","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ef9a5b1edd231d21981de6014ce26e9d397a4cf791d777951a6d9aff455198f6","sha256:6420ba14ded845c8ff2647322a296d7409beb2f48afb88af9a9f03ca7ff47eb1"],"state_sha256":"78a600619132a51e380f6f3e8302e61c0fa35ccd37d0d81826e2b49c6b907c44"}