When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMs

TL;DR

This paper reveals that chain-of-thought prompting can impair instruction-following accuracy in large language models, and proposes strategies to mitigate these negative effects, highlighting a previously overlooked pitfall in reasoning-enhanced LLMs.

Contribution

It systematically uncovers the negative impact of explicit reasoning on instruction-following and introduces practical mitigation techniques, including a novel attention-based metric.

Findings

CoT prompting often reduces instruction-following accuracy.

Selective reasoning strategies can recover performance.

Attention diversion explains reasoning-induced failures.

Abstract

Reasoning-enhanced large language models (RLLMs), whether explicitly trained for reasoning or prompted via chain-of-thought (CoT), have achieved state-of-the-art performance on many complex reasoning tasks. However, we uncover a surprising and previously overlooked phenomenon: explicit CoT reasoning can significantly degrade instruction-following accuracy. Evaluating 15 models on two benchmarks: IFEval (with simple, rule-verifiable constraints) and ComplexBench (with complex, compositional constraints), we consistently observe performance drops when CoT prompting is applied. Through large-scale case studies and an attention-based analysis, we identify common patterns where reasoning either helps (e.g., with formatting or lexical precision) or hurts (e.g., by neglecting simple constraints or introducing unnecessary content). We propose a metric, constraint attention, to quantify model focus during generation and show that CoT reasoning often diverts attention away from instruction-relevant tokens. To mitigate these effects, we introduce and evaluate four strategies: in-context learning, self-reflection, self-selective reasoning, and classifier-selective reasoning. Our results demonstrate that selective reasoning strategies, particularly classifier-selective reasoning, can substantially recover lost performance. To our knowledge, this is the first work to systematically expose reasoning-induced failures in instruction-following and offer practical mitigation strategies.