When Is Voting Enough? Code and model outputs

Code and model outputs for the article "When Is Voting Enough? Fusing the Judgements of Large Language Models on Single and Set-Valued Answers". The archive contains the code, the scripts that download the benchmarks from their original sources and check them against the copies used in the article, the data splits, the parsed and raw outputs of all model calls (six open-weight models) and the evaluation results. The benchmark texts are not redistributed. See README.md for instructions to reproduce every table and figure.