Random Attribution
Sanity-check floor expected to perform near chance across all tasks.
EvalAI-hosted rankings with track filters and score breakdowns.
Sanity-check floor expected to perform near chance across all tasks.
Simple baseline that validates schema and submission pipeline behavior.
Stronger baseline for competitive public-track performance.