Code Simulation

Code simulation research focuses on using computational models, particularly large language models (LLMs) and reinforcement learning (RL) agents, to predict the behavior of code without direct execution. Current efforts concentrate on improving the accuracy and efficiency of LLMs in simulating various code types, from simple programs to complex circuits, often employing novel prompting techniques like "Chain of Simulation" to enhance performance and mitigate limitations like reliance on pattern recognition. This research is significant for improving software testing, hardware verification (e.g., detecting hardware Trojans), and optimizing circuit design through faster power estimation, ultimately leading to more robust and efficient systems.

Papers

September 11, 2024

CPSample: Classifier Protected Sampling for Guarding Training Data During Diffusion
Joshua Kazdan, Hao Sun, Jiaqi Han, Felix Petersen, Stefano Ermon
Training Data Diffusion Explainer Training Image Rejection Sampling Code Simulation

March 24, 2024

Can Language Models Pretend Solvers? Logic Code Simulation with LLMs
Minyu Chen, Guoqiang Li, Ling-I Wu, Ruibang Liu, Yuxin Su, Xi Chang, Jianxin Xue
Large Language Model Language Model Automatic Solver LLM Based Code Generation Natural Language Reasoning Problem Code Simulation

January 17, 2024

Code Simulation Challenges for Large Language Models
Emanuele La Malfa, Christoph Weinhuber, Orazio Torre, Fangru Lin, Samuele Marro, Anthony Cohn, Nigel Shadbolt, Michael Wooldridge
Reasoning Task Algorithmic Reasoning Algorithmic Task Algorithmic Perspective Code Simulation

August 26, 2022

DETERRENT: Detecting Trojans using Reinforcement Learning
Vasudev Gohil, Satwik Patnaik, Hao Guo, Dileep Kalathil, Jeyavijayan, Rajendran
Reinforcement Learning Reinforcement Learning Agent Trojan Detection Hardware Trojan Code Simulation

March 11, 2022

GATSPI: GPU Accelerated Gate-Level Simulation for Power Improvement
Yanqing Zhang, Haoxing Ren, Akshay Sridharan, Brucek Khailany
Performance Improvement Fast Simulation Hardware Simulation Power Prediction Code Simulation

Code Simulation

Papers

CPSample: Classifier Protected Sampling for Guarding Training Data During Diffusion

Can Language Models Pretend Solvers? Logic Code Simulation with LLMs

Code Simulation Challenges for Large Language Models

DETERRENT: Detecting Trojans using Reinforcement Learning

GATSPI: GPU Accelerated Gate-Level Simulation for Power Improvement