Answer-Centric or Reasoning-Driven? Uncovering the Latent Memory Anchor in LLMs

TL;DR

This paper investigates whether LLMs rely more on explicit answers or reasoning patterns, revealing a strong dependence on answer cues and questioning the true inferential capabilities of these models.

Contribution

It introduces a five-level answer-visibility prompt framework and provides empirical evidence that LLMs predominantly depend on answer cues rather than genuine reasoning.

Findings

Performance drops by 26.90% when answer cues are masked

LLMs rely heavily on explicit answers over reasoning chains

Much of LLM reasoning may be post-hoc rationalization

Abstract

While Large Language Models (LLMs) demonstrate impressive reasoning capabilities, growing evidence suggests much of their success stems from memorized answer-reasoning patterns rather than genuine inference. In this work, we investigate a central question: are LLMs primarily anchored to final answers or to the textual pattern of reasoning chains? We propose a five-level answer-visibility prompt framework that systematically manipulates answer cues and probes model behavior through indirect, behavioral analysis. Experiments across state-of-the-art LLMs reveal a strong and consistent reliance on explicit answers. The performance drops by 26.90\% when answer cues are masked, even with complete reasoning chains. These findings suggest that much of the reasoning exhibited by LLMs may reflect post-hoc rationalization rather than true inference, calling into question their inferential depth. Our study uncovers the answer-anchoring phenomenon with rigorous empirical validation and underscores the need for a more nuanced understanding of what constitutes reasoning in LLMs.