The Zero-Step Thinking: An Empirical Study of Mode Selection as Harder Early Exit in Reasoning Models

2025-10-21

AI Summary

Large language models (LLMs) often incur redundant computation in mathematical and logical reasoning due to excessive chain-of-thought (CoT) reasoning. Existing early-exit mechanisms rely on intermediate states, limiting efficiency and adaptability. Method: This paper introduces “pattern selection”—a novel paradigm that determines *a priori*, from the input alone (i.e., zero-step reasoning), whether to deploy a long-chain (Long-CoT) or short-chain (Short-CoT) strategy. We formulate this as an information-constrained early-exit variant and propose a dual-path discriminative framework: one leveraging prompt engineering, the other exploiting internal model representations (e.g., attention weights and hidden states). Contribution/Results: Evaluated across nine benchmarks, methods utilizing internal representations achieve significantly higher classification accuracy than output-only approaches; pure output-driven strategies exhibit limited performance. Our analysis further exposes the fundamental limitations of zero-step decision-making under severe information scarcity, revealing intrinsic trade-offs between efficiency and reliability in adaptive reasoning.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Natural Language Processing: (Large) Language Models

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Reasoning models have demonstrated exceptional performance in tasks such as mathematics and logical reasoning, primarily due to their ability to engage in step-by-step thinking during the reasoning process. However, this often leads to overthinking, resulting in unnecessary computational overhead. To address this issue, Mode Selection aims to automatically decide between Long-CoT (Chain-of-Thought) or Short-CoT by utilizing either a Thinking or NoThinking mode. Simultaneously, Early Exit determines the optimal stopping point during the iterative reasoning process. Both methods seek to reduce the computational burden. In this paper, we first identify Mode Selection as a more challenging variant of the Early Exit problem, as they share similar objectives but differ in decision timing. While Early Exit focuses on determining the best stopping point for concise reasoning at inference time, Mode Selection must make this decision at the beginning of the reasoning process, relying on pre-defined fake thoughts without engaging in an explicit reasoning process, referred to as zero-step thinking. Through empirical studies on nine baselines, we observe that prompt-based approaches often fail due to their limited classification capabilities when provided with minimal hand-crafted information. In contrast, approaches that leverage internal information generally perform better across most scenarios but still exhibit issues with stability. Our findings indicate that existing methods relying solely on the information provided by models are insufficient for effectively addressing Mode Selection in scenarios with limited information, highlighting the ongoing challenges of this task. Our code is available at https://github.com/Trae1ounG/Zero_Step_Thinking.

Problem

• Mode Selection determines reasoning approach before process begins
• Early Exit optimizes computational cost by stopping reasoning early
• Zero-step thinking uses predefined thoughts without explicit reasoning steps

Innovation

• Mode Selection decides Long-CoT or Short-CoT automatically
• Early Exit determines optimal stopping point in reasoning
• Zero-step thinking uses pre-defined fake thoughts without reasoning