Reproduce “LLM-Assisted Optimization for Simulation-Based Redesign” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chan-Eui Song, Jeremy J. Michalek, Christopher McComb
WAFER tests whether large language models can guide simulator-based redesign using changed requirements and an existing design. Across four BatPaC scenarios, LLM-assisted methods match numerical baselines’ feasible-discovery rate after four simulator evaluations, compared with 20. Feasible designs within 2% of best observed cost occur in 23% of LLM-assisted runs and 7% of numerical runs; this does not establish global optimality or faster elapsed time.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.