ResearchBenchmarksCoffee Task on DexMimicGen (test)Follow51Success RateWARP6.2817.8929.541.11Jun 29, 2026Evaluation ResultsMethodMethodLinksSuccess RateWARPEvaluation Protocol=re...Evaluation Protocol=replay2026.0651MINKEvaluation Protocol=re...Evaluation Protocol=replay2026.0650.5WARPEvaluation Protocol=po...Evaluation Protocol=policy rollout2026.0634MINKEvaluation Protocol=po...Evaluation Protocol=policy rollout2026.068