Anti Guessing Olympiad: Algorithmic Math Benchmark for Robust Evaluation

An algorithmic anti-guessing pipeline constructed this olympiad math benchmark to be robust against memorization. The dataset was created by anonymous-antiguessing-2026 and released on Hugging Face, with a last update timestamp of 2026-05-07 00:12:09. It ships with three complementary subsets to support headline evaluation.

Use Cases

• Evaluating model robustness to memorization based on the per-problem verification gap metric.
• Benchmarking solution-correctness versus answer-only accuracy as described in the construction pipeline.
• Training models on olympiad-level math problems designed to resist pattern guessing.

Strengths

• Designed via a three-power-tier algorithmic pipeline to be memorization-robust.
• Operationalizes a specific verification gap as both a construction criterion and evaluation metric.
• Includes three complementary subsets for evaluation, as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

anonymous-antiguessing-2026 on Hugging Face

Collection Method

Constructed via an algorithmic anti-guessing pipeline.

Time Range

The benchmark's temporal coverage is not specified.

Freshness

Last updated 2026-05 07 00:12:09; freshness should be verified.

Geography

The dataset's spatial coverage is not specified.