ResearchBenchmarksAutonomous Agent Task Solving on AgentBoard PDDLFollow76.7Task Completion RateAdMem31.56443.2825566.718Jun 5, 2026Evaluation ResultsMethodMethodLinksTask Completion RateAverage Progress ScoreAdMemLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0676.790.69AWMLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0643.364.07ReActLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0633.355.37