Fig. 3: Graph traversal tasks and results.
From: A brain-inspired agentic architecture to improve planning with LLMs
A Graph traversal tasks. Steppath: Agent must identify the shortest path from a start state to a goal state. Valuepath: Agent must identify the shortest path from a start state to the state with the largest reward, while avoiding the state with the smaller reward. B Graph traversal results. `% solved' indicates the percentage of problems solved without proposing invalid actions (↑ better). GPT-4 Zero-shot, ICL, COT, and MAD baselines are deterministic, and therefore, a single run was performed on all problems. Note that MAP did not employ tree search on the Steppath task, and did not employ task decomposition on any of the graph traversal tasks. Without tree search, MAP's performance is deterministic, and therefore only a single run was performed on the Steppath task, whereas we performed 5 runs with ToT. Gray error bars reflect 95% binomial confidence intervals (for models evaluated on a single run). Black dots indicate performance for individual runs. Colored dots reflect values of 0%. Dark bars indicate average performance over multiple plans/runs. Light bars indicate best performance. For Valuepath, Detour, and Reward Revaluation, we performed 10, 10, and 5 runs, respectively, with MAP and ToT, and present average performance ± the standard error of the mean (black error bars). See Supplementary Section S3 for results in tabular form.
