Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

One-step next-observation prediction research benchmarks · SOTA2 Research

One-step next-observation prediction

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

AgentGym Unweighted Average (test)

85Token F1

16

Jun 1, 2026

PatchWorld-Residual

0.72Token F1

16

Jun 1, 2026

63Token F1

16

Jun 1, 2026

TextCraft (test)

PatchWorld-Residual

95Token F1

16

Jun 1, 2026

SciWorld (test)

96Token F1

16

Jun 1, 2026

PatchWorld-Residual

98Token F1

16

Jun 1, 2026

93Token F1

16

Jun 1, 2026

AlfWorld (test)

89Token F1

16

Jun 1, 2026

Popular datasets

AlfWorld MAZE BabyAI Webshop Sciworld

Follow for updates

Join SOTA2 Contact