Code Simulation Challenges for Large Language Models

TL;DR

This paper evaluates the ability of Large Language Models to simulate coding and algorithmic tasks, introduces benchmarks, and proposes a novel prompting method called Chain of Simulation (CoSm) to improve their reasoning capabilities.

Contribution

It introduces new benchmarks for code simulation and a novel prompting technique, CoSm, to enhance LLMs' algorithmic reasoning and simulation performance.

Findings

LLMs' simulation ability is affected by algorithmic complexity.

Powerful LLMs show relatively strong but fragile simulation capabilities.

CoSm improves simulation performance by reducing reliance on memorization.

Abstract

Many reasoning, planning, and problem-solving tasks share an intrinsic algorithmic nature: correctly simulating each step is a sufficient condition to solve them correctly. This work studies to what extent Large Language Models (LLMs) can simulate coding and algorithmic tasks to provide insights into general capabilities in such algorithmic reasoning tasks. We introduce benchmarks for straight-line programs, code that contains critical paths, and approximate and redundant instructions. We further assess the simulation capabilities of LLMs with sorting algorithms and nested loops and show that a routine's computational complexity directly affects an LLM's ability to simulate its execution. While the most powerful LLMs exhibit relatively strong simulation capabilities, the process is fragile, seems to rely heavily on pattern recognition, and is affected by memorisation. We propose a novel off-the-shelf prompting method, Chain of Simulation (CoSm), which instructs LLMs to simulate code execution line by line/follow the computation pattern of compilers. CoSm efficiently helps LLMs reduce memorisation and shallow pattern recognition while improving simulation performance. We consider the success of CoSm in code simulation to be inspirational for other general routine simulation reasoning tasks.