What it measures

Knowledge-Orthogonal Reasoning benchmark testing rule application over newly invented symbolic, cipher, puzzle, and counterfactual rules.

Reading the measurement

Study what information the model receives and which inference the answer requires. A question may mix recall, visual perception, deduction and search. Use task breakdowns to find the mechanism behind a headline gain.

Published evaluation

Overall accuracy; KOR-Bench · main comparison. The source’s task split, model configuration and scoring units are retained.

Where this fits in an evaluation suite

Ask what the task actually requires: recall, deduction, search, or a mixture. Separate closed-book reasoning from retrieval and tool-assisted problem solving.

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

Overall accuracy; KOR-Bench · main comparison. The source’s task split, model configuration and scoring units are retained.

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

  1. KOR-Bench: Benchmarking Language Models on Knowledge-Orthogonal Reasoning Tasks — published result table ↗

    Kaijing Ma, Xinrun Du, Yunran Wang, Haoran Zhang, Zhoufutu Wen, Xingwei Qu et al. · 2024

    All 12 authors

    Kaijing Ma, Xinrun Du, Yunran Wang, Haoran Zhang, Zhoufutu Wen, Xingwei Qu, Jian Yang, Jiaheng Liu, Minghao Liu, Xiang Yue, Wenhao Huang, Ge Zhang

Reviewed 2026-10-26. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.