Chain of Simulation: A Dual-Mode Reasoning Framework for Large Language Models with Dynamic Problem Routing

We present Chain of Simulation (CoS), a novel dual-mode reasoning framework that dynamically routes problems to specialized reasoning strategies in Large Language Models (LLMs). Unlike existing uniform prompting approaches, CoS employs three distinct reasoning modes: (1) computational flow with self-consistency for mathematical problems, (2) symbolic state tracking with JSON representations for spatial reasoning, and (3) hybrid fact-extraction for multi-hop inference. Through comprehensive evaluation on GSM8K, StrategyQA, and bAbI benchmarks using four state-of-the-art models (Gemma-3 27B, LLaMA-3.1 8B, Mistral 7B, and Qwen-2.5 14B), we demonstrate that CoS achieves 71.5% accuracy on GSM8K (1.0% absolute improvement), 90.0% on StrategyQA (2.5% improvement), and 19.0% on bAbI (65.2% relative improvement) compared to the strongest baselines. The analysis reveals that problem-specific mode selection is crucial, with computational mode achieving 81.2% accuracy when correctly applied to mathematical problems, while misrouting leads to 0% accuracy. We provide detailed algorithms for mode selection, state tracking, and answer extraction, establishing CoS as an effective approach for improving LLM reasoning without additional training. The framework provides superior trade-offs between accuracy and efficiency compared to Self-Consistency, achieving comparable performance at 54% lower computational cost.

💡 Research Summary

The paper introduces Chain of Simulation (CoS), a dual‑mode reasoning framework that dynamically routes each input problem to the most suitable reasoning strategy within a frozen large language model (LLM). The authors argue that the prevailing uniform prompting approaches—such as standard Chain‑of‑Thought (CoT) or Self‑Consistency—apply the same reasoning pattern regardless of problem type, thereby under‑utilizing latent capabilities of the model. Inspired by dual‑process theory in cognitive psychology, CoS treats a pre‑trained LLM as possessing multiple “sub‑systems” that can be selectively activated through carefully crafted prompts.

CoS consists of four sequential components:

1. Problem Analyzer – The input text (context + question) is lower‑cased, then processed with regular expressions and keyword lists to produce a binary feature vector a =