ResearchBenchmarksInteractive Learning on Square NoisyFollow94.6Success RateSDP-3.78421.75847.372.842Jun 1, 2026Evaluation ResultsMethodMethodLinksSuccess RateSDPFeedback=NoisyFeedback=Noisy2026.0694.6ADPFeedback=NoisyFeedback=Noisy2026.0683.8DPFeedback=NoisyFeedback=Noisy2026.0668.3DP-DPOFeedback=NoisyFeedback=Noisy2026.0662.3CLICFeedback=NoisyFeedback=Noisy2026.0642.9IBCFeedback=NoisyFeedback=Noisy2026.060