ResearchBenchmarksAutomated agent rollback on GSM8K Right Dense variant (test)Follow75Success RateLF69.871.1572.573.85May 7, 2026Evaluation ResultsMethodMethodLinksSuccess RateCoverageCostLFScoring function=fine-...Scoring function=fine-tuned LLM, Target coverage=80%2026.0575820.4Top-1Scoring function=fine-...Scoring function=fine-tuned LLM2026.0571860.5VCPScoring function=fine-...Scoring function=fine-tuned LLM, Target coverage=80%2026.05701000.66