Figure 2. Operational workflow used in this study to assess practical reproducibility on MatBench. LLM: Large language model.