# Judge Agreement Accuracy on Zhihu 847 queries (test) Source: [Judge Agreement Accuracy on Zhihu 847 queries (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:34:46.735Z Evaluation of judge agreement accuracy using the test split of the Zhihu 847 queries dataset. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Judge Agreement Accuracy | | Dataset (including any stated split) | Zhihu 847 queries (test) | | Metric | Agreement Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Agreement Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Agreement Accuracy | Result date | Report paper | | --- | --- | --- | --- | | LogicJudge | 0.75 | Not recorded | 699d30576017c4460ffab1d5 | | gemini-2.5-pro | 0.74 | Not recorded | 699d30576017c4460ffab1d5 | | Ensemble Vote | 0.7357 | Not recorded | 699d30576017c4460ffab1d5 | | gpt-o3 | 0.717 | Not recorded | 699d30576017c4460ffab1d5 | | qwen3-max | 0.717 | Not recorded | 699d30576017c4460ffab1d5 | | gpt-5 | 0.716 | Not recorded | 699d30576017c4460ffab1d5 | | gpt-4.1 | 0.7077 | Not recorded | 699d30576017c4460ffab1d5 | | gemini-3-pro | 0.696 | Not recorded | 699d30576017c4460ffab1d5 | | gemini-3-pro | 0.682 | Not recorded | 699d30576017c4460ffab1d5 | | deepseek-v3 | 0.6641 | Not recorded | 699d30576017c4460ffab1d5 | | gpt-5.1 | 0.65 | Not recorded | 699d30576017c4460ffab1d5 | | claude-4.5-sonnet | 0.645 | Not recorded | 699d30576017c4460ffab1d5 | | claude-4.5-sonnet | 0.64 | Not recorded | 699d30576017c4460ffab1d5 | | qwen-max | 0.6162 | Not recorded | 699d30576017c4460ffab1d5 | | Ensemble Consensus | 0.6111 | Not recorded | 699d30576017c4460ffab1d5 | | claude-4-sonnet | 0.567 | Not recorded | 699d30576017c4460ffab1d5 | | claude-4-sonnet | 0.562 | Not recorded | 699d30576017c4460ffab1d5 | | qwen3-235b | 0.492 | Not recorded | 699d30576017c4460ffab1d5 | | gemini-2.5-pro | 0.419 | Not recorded | 699d30576017c4460ffab1d5 | ### LogicJudge Agreement Accuracy: 0.75. Result ID: 699e14275698665ec4cc791d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"backbone":"Qwen-3-30B-A3B","training":"SFT + GRPO"}. This record does not assert independent reproduction. ### gemini-2.5-pro Agreement Accuracy: 0.74. Result ID: 699e14275698665ec4cc790f. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### Ensemble Vote Agreement Accuracy: 0.7357. Result ID: 699e14275698665ec4cc791c. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gpt-o3 Agreement Accuracy: 0.717. Result ID: 699e14275698665ec4cc790b. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### qwen3-max Agreement Accuracy: 0.717. Result ID: 699e14275698665ec4cc7917. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gpt-5 Agreement Accuracy: 0.716. Result ID: 699e14275698665ec4cc790d. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gpt-4.1 Agreement Accuracy: 0.7077. Result ID: 699e14275698665ec4cc790c. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gemini-3-pro Agreement Accuracy: 0.696. Result ID: 699e14275698665ec4cc7912. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot","mode":"thinking"}. This record does not assert independent reproduction. ### gemini-3-pro Agreement Accuracy: 0.682. Result ID: 699e14275698665ec4cc7911. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### deepseek-v3 Agreement Accuracy: 0.6641. Result ID: 699e14275698665ec4cc791a. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gpt-5.1 Agreement Accuracy: 0.65. Result ID: 699e14275698665ec4cc790e. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### claude-4.5-sonnet Agreement Accuracy: 0.645. Result ID: 699e14275698665ec4cc7916. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot","mode":"thinking"}. This record does not assert independent reproduction. ### claude-4.5-sonnet Agreement Accuracy: 0.64. Result ID: 699e14275698665ec4cc7915. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### qwen-max Agreement Accuracy: 0.6162. Result ID: 699e14275698665ec4cc7918. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### Ensemble Consensus Agreement Accuracy: 0.6111. Result ID: 699e14275698665ec4cc791b. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### claude-4-sonnet Agreement Accuracy: 0.567. Result ID: 699e14275698665ec4cc7913. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### claude-4-sonnet Agreement Accuracy: 0.562. Result ID: 699e14275698665ec4cc7914. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot","mode":"thinking"}. This record does not assert independent reproduction. ### qwen3-235b Agreement Accuracy: 0.492. Result ID: 699e14275698665ec4cc7919. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot"}. This record does not assert independent reproduction. ### gemini-2.5-pro Agreement Accuracy: 0.419. Result ID: 699e14275698665ec4cc7910. [Reporting paper]() Recorded evaluation configuration: {"evaluation\_setting":"one-shot","mode":"thinking"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.