This dataset is the released sweep data for the shared task at the COLM 2026
Workshop on LLM-Based Social Simulations. The task is to build evaluation
methods over these simulation logs. Each study ships hypotheses you can
evaluate (does the data support or refute them?), and you may also pose and test
your own questions grounded in the scenario and its data. Inventing the
operationalization is the task;… See the full description on the dataset page:
https://huggingface.co/datasets/ComplexDataLab/socsim26-sharedtask.
