Chain of Simulation: A Dual-Mode Reasoning Framework for Large Language Models with Dynamic Problem Routing
Saeid Sheikhi · 2026
Read Chain of Simulation: A Dual-Mode Reasoning Framework for Large Language Models with Dynamic Problem Routing by Saeid Sheikhi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-02842.txt.
Description
Routing each problem to the right reasoning style is the idea in Saeid Sheikhi's solo 2026 paper. His Chain of Simulation framework gives language models three modes. One handles math with calculation and self-checking. Another tracks positions using structured data. A third pulls facts together for multi-step questions.
Sheikhi tests the system on math, yes-or-no reasoning, and story-based benchmarks. Four different language models are used. Routing improved accuracy on all three benchmarks over the strongest baselines. The analysis also shows a sharp warning. Sending a problem to the wrong mode can drop accuracy to zero, so correct routing matters as much as the modes themselves.
Catalog record
| CREATOR | Saeid Sheikhi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-02842