Automated LLM-based evaluation of code review bot comments achieves only moderate agreement (0.44-0.62) with developer labels in an industrial dataset because developer decisions reflect contextual constraints beyond comment quality.
Think again! the effect of test-time compute on preferences, opinions, and beliefs of large language models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Understanding the Limits of Automated Evaluation for Code Review Bots in Practice
Automated LLM-based evaluation of code review bot comments achieves only moderate agreement (0.44-0.62) with developer labels in an industrial dataset because developer decisions reflect contextual constraints beyond comment quality.