Large Language Model Powered Selective Concolic Testing

This is the artifact of the paper **"Large Language Model Powered Selective Concolic Testing"** . The artifact is provided as a Docker image, which contains the prototype of the proposed method and all 4 floating-point benchmarks used in the evaluation. The aim is to assist the reviewers in reproducing the experimental results reported in the paper. The complete running logs and intermediate results are saved in **sctllm-results-0815.tar.gz**. If you only focus on the intermediate results, you can directly decompress the compressed file and run `python3 export_all.py` to export all the data and figures required in the paper.