Merlin's Whisper: Enabling Efficient Reasoning in Large Language Models via Black-box Persuasive Prompting
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911691318493184 |
|---|---|
| author | Xia, Heming Du, Cunxiao Li, Rui Leong, Chak Tou Li, Yongqi Li, Wenjie |
| author_facet | Xia, Heming Du, Cunxiao Li, Rui Leong, Chak Tou Li, Yongqi Li, Wenjie |
| contents | Large reasoning models (LRMs) have demonstrated remarkable proficiency in tackling complex tasks through step-by-step thinking. However, this lengthy reasoning process incurs substantial computational and latency overheads, hindering the practical deployment of LRMs. This work presents a new approach to mitigating overthinking in LRMs via black-box persuasive prompting. By treating LRMs as black-box communicators, we investigate how to persuade them to generate concise responses without compromising accuracy. We introduce Whisper, an iterative refinement framework that generates high-quality persuasive prompts from diverse perspectives. Experiments across multiple benchmarks demonstrate that Whisper consistently reduces token usage while preserving performance. Notably, Whisper 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 all benchmarks. For closed-source APIs, Whisper reduces token usage on MATH-500 by 46% for Claude-3.7 and 50% for Gemini-2.5. Further analysis reveals the broad applicability of Whisper across data domains, model scales, and families, underscoring the potential of black-box persuasive prompting as a practical strategy for enhancing LRM efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_10528 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Merlin's Whisper: Enabling Efficient Reasoning in Large Language Models via Black-box Persuasive Prompting Xia, Heming Du, Cunxiao Li, Rui Leong, Chak Tou Li, Yongqi Li, Wenjie Computation and Language Machine Learning Large reasoning models (LRMs) have demonstrated remarkable proficiency in tackling complex tasks through step-by-step thinking. However, this lengthy reasoning process incurs substantial computational and latency overheads, hindering the practical deployment of LRMs. This work presents a new approach to mitigating overthinking in LRMs via black-box persuasive prompting. By treating LRMs as black-box communicators, we investigate how to persuade them to generate concise responses without compromising accuracy. We introduce Whisper, an iterative refinement framework that generates high-quality persuasive prompts from diverse perspectives. Experiments across multiple benchmarks demonstrate that Whisper consistently reduces token usage while preserving performance. Notably, Whisper 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 all benchmarks. For closed-source APIs, Whisper reduces token usage on MATH-500 by 46% for Claude-3.7 and 50% for Gemini-2.5. Further analysis reveals the broad applicability of Whisper across data domains, model scales, and families, underscoring the potential of black-box persuasive prompting as a practical strategy for enhancing LRM efficiency. |
| title | Merlin's Whisper: Enabling Efficient Reasoning in Large Language Models via Black-box Persuasive Prompting |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2510.10528 |