# Diversity of solution strategies on AIME 2024 + 2025 Source: [Diversity of solution strategies on AIME 2024 + 2025]() Collection index: [llms.txt]() Data updated: 2026-05-19T09:52:02.598Z Evaluation of model performance focusing on the diversity and variability of solution strategies generated for the combined AIME 2024 and 2025 mathematics competitions. 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 | Diversity of solution strategies | | Dataset (including any stated split) | AIME 2024 + 2025 | | Metric | Avg Diversity | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 5 of 5 | | 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 - [Avg Diversity](): Higher is better; unit: SCALAR. - [Std Dev](): Higher is better; unit: SCALAR. - [Diversity (AIME 2024)](): Higher is better; unit: SCALAR. - [Diversity (AIME 2025)](): Higher is better; unit: SCALAR. ## Recorded results | Method | Avg Diversity | Result date | Report paper | | --- | --- | --- | --- | | Qwen2.5-7B | 3.9 | Not recorded | 6a0bee98dbc836b87d82c430 | | Qwen2.5-7B + DISA | 3.72 | Not recorded | 6a0bee98dbc836b87d82c430 | | Qwen2.5-7B + FlowRL | 3.4 | Not recorded | 6a0bee98dbc836b87d82c430 | | Qwen2.5-7B + GRPO | 3.02 | Not recorded | 6a0bee98dbc836b87d82c430 | | Qwen2.5-7B + GSPO | 2.97 | Not recorded | 6a0bee98dbc836b87d82c430 | ### Qwen2.5-7B Avg Diversity: 3.9. Result ID: 6a0c314afbc373592c02b70d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method":"Backbone"}. This record does not assert independent reproduction. ### Qwen2.5-7B + DISA Avg Diversity: 3.72. Result ID: 6a0c314afbc373592c02b711. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method":"RL-post-trained variant"}. This record does not assert independent reproduction. ### Qwen2.5-7B + FlowRL Avg Diversity: 3.4. Result ID: 6a0c314afbc373592c02b710. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method":"RL-post-trained variant"}. This record does not assert independent reproduction. ### Qwen2.5-7B + GRPO Avg Diversity: 3.02. Result ID: 6a0c314afbc373592c02b70e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Method":"RL-post-trained variant"}. This record does not assert independent reproduction. ### Qwen2.5-7B + GSPO Avg Diversity: 2.97. Result ID: 6a0c314afbc373592c02b70f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method":"RL-post-trained variant"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.