ResearchBenchmarksList operations evaluation on ListOps (3, 14) (test)Follow79.1Mean AccuracyRIR-GRC23.35637.82852.366.772May 25, 2026Evaluation ResultsMethodMethodLinksMean AccuracyLongest BinRIR-GRC2026.0579.166.4BBT-GRC2026.0579.167.7MLP-LDRU2026.0569.765.2TF (ALiBi)Positional Encoding=ALiBiPositional Encoding=ALiBi2026.0563.346.4LSTM2026.0562.751.4TF (NoPE)Positional Encoding=NonePositional Encoding=None2026.0537.631.2TF (Sin.)Positional Encoding=Si...Positional Encoding=Sinusoidal2026.0525.58.4