ResearchBenchmarksAutonomous Agent Task Solving on AgentBoard JerichoFollow50Task CompletenessAdMem8.419.23040.8Jun 5, 2026Evaluation ResultsMethodMethodLinksTask CompletenessAverage ProgressAdMemLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.065053.25ReActLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.064040.02AWMLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.061022.27