What it measures
Contest mathematics with short integer answers provides a compact test of multi-step problem solving.
How to read the result
Report the number of attempts and actual correct count alongside a percentage.
Limits and comparison traps
The set is small. Answer correctness alone does not distinguish a valid derivation from a lucky or memorized answer.
Benchmark reference notes
30 problems from the 2025 American Invitational Mathematics Examination (I & II).
Where this fits in an evaluation suite
Final-answer accuracy, proof credit, and expert-reviewed research progress are distinct signals. A proof-aware evaluation should retain partial work and audit strong claims.
Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.
A reproducible comparison
Use the exact implementation, version, task subset, tool access, attempt budget, and judge described by the source.
Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.
PRIMARY EVIDENCE & FURTHER READING
Reviewed 2026-10-26. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.