Merlin's Whisper: Enabling Efficient Reasoning in LLMs via Black-box Adversarial Prompting
This work addresses the practical deployment challenge of LRMs by improving efficiency, but it is incremental as it builds on existing prompting methods.
The paper tackles the problem of computational inefficiency in large reasoning models (LRMs) due to lengthy reasoning processes, and introduces AdvPrompt, a black-box adversarial prompting framework that reduces token usage while preserving accuracy, achieving up to a 3x reduction in response length on GSM8K and 35-47% token reduction on MATH-500 for various models.
Large reasoning models (LRMs) have demonstrated remarkable proficiency in tackling complex reasoning tasks through step-by-step thinking. However, such a lengthy reasoning process incurs substantial computational and latency overheads, hindering the practical deployment of these models. In this work, we present a new perspective on mitigating overthinking in LRMs via black-box adversarial prompting. By treating both open-source LRMs and closed-source APIs as black-box communicators, we investigate how to elicit concise responses without sacrificing accuracy. We introduce AdvPrompt, an iterative refinement framework that generates high-quality adversarial prompts from diverse perspectives. Experiments across multiple benchmarks demonstrate that AdvPrompt consistently reduces token usage while preserving performance. Notably, AdvPrompt achieves a 3x reduction in average response length on simple GSM8K questions for the Qwen3 model series, and delivers an average ~40% token reduction across four benchmarks. For closed-source APIs, AdvPrompt reduces token usage on MATH-500 by 35% for Claude-3.7 and 47% for Gemini-2.5. Further analysis reveals the generalizability of AdvPrompt across various model scales and families, underscoring the potential of black-box prompting as a practical and effective strategy for enhancing LRM efficiency.