ResearchBenchmarksInteractive Learning on Push-T NoisyFollow60Success RateSDP32.23239.44146.6553.859Jun 1, 2026Evaluation ResultsMethodMethodLinksSuccess RateSDPFeedback=NoisyFeedback=Noisy2026.0660ADPFeedback=NoisyFeedback=Noisy2026.0641DP-DPOFeedback=NoisyFeedback=Noisy2026.0640CLICFeedback=NoisyFeedback=Noisy2026.0639DPFeedback=NoisyFeedback=Noisy2026.0633.3IBCFeedback=NoisyFeedback=Noisy2026.0633.3