CAP-CoT: Cycle Adversarial Prompt for Improving Chain of Thoughts in LLM Reasoning
For practitioners deploying LLMs on multi-step reasoning tasks, CAP-CoT offers a method to enhance reasoning consistency and robustness without requiring model retraining.
CAP-CoT introduces a cycle adversarial prompt optimization framework that improves both accuracy and stability of Chain-of-Thought reasoning in LLMs by iteratively contrasting correct and deliberately flawed reasoning chains. Across six benchmarks and four LLM backbones, it reduces variability and improves accuracy within 2-3 cycles.
Chain-of-Thought (CoT) prompting has emerged as a simple and effective way to elicit step-by-step solutions from large language models (LLMs). However, CoT reasoning can be unstable across runs on long, multi-step problems, leading to inconsistent answers for unchanged task. Most prior work focuses on improving the forward reasoning chain within a single pass, with less attention to iterative and contrastive correction. To address this gap, we propose CAP-CoT, a Cycle Adversarial Prompt optimization framework designed to improve both CoT reasoning accuracy and stability of a single deployed solver. In each cycle, a forward solver generates candidate reasoning chains, an adversarial challenger constructs plausible but deliberately flawed chains using targeted error strategies, and a feedback agent contrasts the two chains and produces step-aligned structured feedback. This feedback closes the optimization loop in two directions, including updating the solver prompt based on errors exposed by the challenger, and updating the challenger prompt to generate increasingly targeted errors in subsequent cycles. Unlike safety-oriented adversarial prompting such as jailbreak or prompt-injection attacks, our adversarial component is task-semantic and aims to expose logical vulnerabilities in reasoning chains. Experiments across six benchmarks and four LLM backbones demonstrate that within two to three adversarial prompt optimization cycles, CAP-CoT consistently reduces variability across runs while improving reasoning accuracy and robustness to prompt perturbations.