What it measures and how
Use a fixed standard harness, separately from results using adapters optimized by providers.
How this evidence is used
Under observation: the official Terms require written permission for systematic collection and aggregation. Aggregation terms for current official results remain unresolved. Standard harness and Provider Adapter results stay separate; the best specialized adapter score does not represent a general model.
Limitations and attribution
Under observation: the official Terms require written permission for systematic collection and aggregation. Aggregation terms for current official results remain unresolved. Standard harness and Provider Adapter results stay separate; the best specialized adapter score does not represent a general model.
Data license: Leaderboard data usage terms remain to be confirmed.
Results are published by ARC Prize. Raw scores and the e/acc consensus index use different scales and cannot be added directly.