---
format: "aidr-story-markdown/v1"
id: "a6c8bece0ccbe43708a7eabbd88d16951c74428c647dc0a19437aea101169efa"
canonical_url: "https://aidr.today/a6c8bece?lang=en"
title: "Thinking Inertia: LLMs Keep Thinking When Told Not To"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-13T00:00:00.000Z"
category: "Research"
topics: ["reasoning"]
source_urls: ["https://huggingface.co/papers/2610.11765"]
summary: "A study examines LLMs' no-thinking behavior using three evaluation metrics: Empty-Thinking Rate, instruction-aware Question-Pre-answer Relevance, and LLM-as-judge Explicit Inference Rate. The research evaluates six prompting interventions on six LLMs across Boolean, multiple-choice, and open-ended questions."
---

# Thinking Inertia: LLMs Keep Thinking When Told Not To

> [Open the canonical story](<https://aidr.today/a6c8bece?lang=en>)

**Published:** 2026-10-13T00:00:00.000Z
**Category:** Research
**Topics:** reasoning

## Summary

A study examines LLMs' no\-thinking behavior using three evaluation metrics: Empty\-Thinking Rate, instruction\-aware Question\-Pre\-answer Relevance, and LLM\-as\-judge Explicit Inference Rate\. The research evaluates six prompting interventions on six LLMs across Boolean, multiple\-choice, and open\-ended questions\.

## Sources

- [Story source](<https://huggingface.co/papers/2610.11765>)