ReviewBench
ReviewBench is built on representative GitHub pull requests ReviewBench is modeled after over 100 million real pull requests on GitHub The article ‘ReviewBench: An open benchmark for AI code review’ appeared first on The GitHub Blog ReviewBench uses Claude Sonnet 5 as the LLM grader to validate findings ReviewBench provides an offline evaluation of Copilot code review (CCR) that is more effective at anticipating the direction of production experiments· Entity summary generated by Graphiti
Relation timeline
BY VALID-FROM DATE- ReviewBenchEVALUATESCopilot code review
ReviewBench provides an offline evaluation of Copilot code review (CCR) that is more effective at anticipating the direction of production experiments
- ReviewBenchBUILT ONGitHub
ReviewBench is built on representative GitHub pull requests
- ReviewBenchUSES GRADERClaude Sonnet 5
ReviewBench uses Claude Sonnet 5 as the LLM grader to validate findings
- GitHub BlogPUBLISHEDReviewBench
The article ‘ReviewBench: An open benchmark for AI code review’ appeared first on The GitHub Blog
- ReviewBenchMODELED AFTERGitHub
ReviewBench is modeled after over 100 million real pull requests on GitHub
Relations are facts Graphiti extracted from reports, not verbatim quotes. An end date means later reporting showed the relation no longer holds.