Meta-Reasoning Improves Tool Use in Large Language Models
This addresses the challenge of improving tool use for large language models in tasks like math reasoning, representing an incremental advance over existing methods.
The paper tackled the problem of suboptimal tool selection in large language models by introducing TECTON, a two-phase system that uses meta-reasoning to choose among candidate tools, resulting in substantial gains on math reasoning datasets both in-distribution and out-of-distribution.
External tools help large language models succeed at tasks where they would otherwise typically fail. In existing frameworks, choosing tools at test time relies on naive greedy decoding, regardless of whether the model has been fine-tuned on tool-annotated data or prompted with in-context examples. In contrast, we find that gathering and choosing among a suitable set of candidate tools has greater potential to lead to an optimal selection. We present Tool selECTion via meta-reasONing (TECTON), a two-phase system that first reasons over a task and outputs candidate tools using a custom fine-tuned language modelling head. Then, with the custom head disabled, it meta-reasons (i.e., it reasons over the previous reasoning process) to make a final choice. We show that TECTON results in substantial gains--both in-distribution and out-of-distribution--on a range of math reasoning datasets.