Trade-offs in Large Reasoning Models: An Empirical Analysis of Deliberative and Adaptive Reasoning over Foundational Capabilities

ACL ARR 2025 May Submission3412 Authors

Abstract: Recent advancements in Large Reasoning Models (LRMs), such as OpenAI's o1/o3 and DeepSeek-R1, have demonstrated remarkable performance in specialized reasoning tasks through human-like deliberative thinking and long chain-of-thought reasoning. However, our systematic evaluation across various model families (DeepSeek, Qwen, and LLaMA) and scales (7B to 671B) reveals that acquiring these deliberative reasoning capabilities significantly reduces the foundational capabilities of LRMs, including notable declines in helpfulness and harmlessness, alongside substantially increased inference costs. Importantly, we demonstrate that adaptive reasoning---employing modes like Zero-Thinking, Less-Thinking, and Summary-Thinking---can effectively alleviate these drawbacks. Our empirical insights underline the critical need for developing more versatile LRMs capable of dynamically allocating inference-time compute according to specific task characteristics.

Paper Type: Long

Research Area: Generation

Research Area Keywords: Large Reasoning Model, Adaptive Reasoning

Contribution Types: NLP engineering experiment

Languages Studied: English

Submission Number: 3412