TORSO: Template-Oriented Reasoning Towards General Tasks
This addresses the cost and inconsistency issues in prompt engineering for LLMs, offering a more generalizable approach for complex problem-solving.
The paper tackles the problem of LLMs' reliance on manually crafted few-shot prompts for reasoning tasks by introducing TORSO, which elicits internal reasoning abilities without such prompts, achieving strong performance on diverse benchmarks.
The approaches that guide Large Language Models (LLMs) to emulate human reasoning during response generation have emerged as an effective method for enabling them to solve complex problems in a step-by-step manner, thereby achieving superior performance. However, most existing approaches using few-shot prompts to generate responses heavily depend on the provided examples, limiting the utilization of the model's inherent reasoning capabilities. Moreover, constructing task-specific few-shot prompts is often costly and may lead to inconsistencies across different tasks. In this work, we introduce Template-Oriented Reasoning (TORSO), which elicits the model to utilize internal reasoning abilities to generate proper responses across various tasks without the need for manually crafted few-shot examples. Our experimental results demonstrate that TORSO achieves strong performance on diverse LLMs benchmarks with reasonable rationales.