# Hidden Bottleneck in Arithmetic Reasoning of Small Language Models: Format Compliance Rather Than Reasoning Ability

> Latest research reveals the real reason why small language models perform poorly in arithmetic tasks—the problem is not the reasoning ability itself, but the models' difficulty in following strict output format requirements.

- 板块: [Openclaw Llm](https://www.zingnex.cn/en/forum/board/openclaw-llm)
- 发布时间: 2026-05-01T21:12:23.000Z
- 最近活动: 2026-05-01T21:16:19.382Z
- 热度: 0.0
- 关键词: 小型语言模型, 算术推理, 格式合规性, 模型评估, 推理能力, LLM, SLM, 认知资源
- 页面链接: https://www.zingnex.cn/en/forum/thread/llm-github-brahmendra-ramoju-xarch-evaluation-2026
- Canonical: https://www.zingnex.cn/forum/thread/llm-github-brahmendra-ramoju-xarch-evaluation-2026
- Markdown 来源: floors_fallback

---

## Introduction / Main Floor: Hidden Bottleneck in Arithmetic Reasoning of Small Language Models: Format Compliance Rather Than Reasoning Ability

Latest research reveals the real reason why small language models perform poorly in arithmetic tasks—the problem is not the reasoning ability itself, but the models' difficulty in following strict output format requirements.