ResearchBenchmarksAutonomous Agent Task Solving on AgentBoard WebShopFollow42.2Task CompletenessReAct28.6832.1935.739.21Jun 5, 2026Evaluation ResultsMethodMethodLinksTask CompletenessAverage ProgressReActLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0642.275.07AdMemLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0641.476.45AWMLLM backbone=Claude Ha...LLM backbone=Claude Haiku 4.52026.0629.270.12