Run evaluations, understand scores, and contribute to TCSExam. Choose a starting point below.
Set up a model, run one task, and inspect the result.
Explore the four tracks and how they contribute to the Index.
Suggest a research problem or help improve an existing task.
- Run more tasks: Lite and Full cover complete-suite commands and Online setup.
- Change a setting: Run configuration covers agents, model access, budgets and result files.
- Prepare your machine: check host requirements and resource preparation.
Browse the task catalogue, compare Solver protocols, or cite the paper.
For service deployment, see the Docker backend guide. For reuse and attribution, see licenses and notices.