PathFinder: Guided Search over Multi-Step Reasoning Paths
This work addresses a key bottleneck in AI reasoning for tasks requiring multiple steps, offering an incremental improvement over existing methods like chain-of-thought prompting.
The paper tackles the challenge of multi-step reasoning in large language models by proposing PathFinder, a tree-search-based approach that enhances reasoning path generation through dynamic decoding and constrained reasoning, achieving an average 6% improvement over baselines on complex arithmetic and commonsense reasoning tasks.
With recent advancements in large language models, methods like chain-of-thought prompting to elicit reasoning chains have been shown to improve results on reasoning tasks. However, tasks that require multiple steps of reasoning still pose significant challenges to state-of-the-art models. Drawing inspiration from the beam search algorithm, we propose PathFinder, a tree-search-based reasoning path generation approach. It enhances diverse branching and multi-hop reasoning through the integration of dynamic decoding, enabled by varying sampling methods and parameters. Using constrained reasoning, PathFinder integrates novel quality constraints, pruning, and exploration methods to enhance the efficiency and the quality of generation. Moreover, it includes scoring and ranking features to improve candidate selection. Our approach outperforms competitive baselines on three complex arithmetic and commonsense reasoning tasks by 6% on average. Our model generalizes well to longer, unseen reasoning chains, reflecting similar complexities to beam search with large branching factors.