# Mathematical Reasoning on Minerva Math (Avg@4, Pass@4) Source: [Mathematical Reasoning on Minerva Math (Avg@4, Pass@4)]() Collection index: [llms.txt]() Data updated: 2026-04-14T11:14:59.573Z Evaluation of model performance on complex mathematical reasoning problems using the Minerva Math dataset, focusing on sampling metrics such as Avg@4 and Pass@4. 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 | Mathematical Reasoning | | Dataset (including any stated split) | Minerva Math | | Metric | Avg@4 | | 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 - [Avg@4](): Higher is better; unit: PERCENT. - [Pass@4](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Avg@4 | Result date | Report paper | | --- | --- | --- | --- | | EAPO | 43.9 | Not recorded | 69ddca697d0172140ec79e8a | | EntroReg | 43.1 | Not recorded | 69ddca697d0172140ec79e8a | | Forking | 43 | Not recorded | 69ddca697d0172140ec79e8a | | DAPO | 42.8 | Not recorded | 69ddca697d0172140ec79e8a | | EntroAdv | 42.7 | Not recorded | 69ddca697d0172140ec79e8a | | KL-Cov | 40.9 | Not recorded | 69ddca697d0172140ec79e8a | | KL-Cov | 40.5 | Not recorded | 69ddca697d0172140ec79e8a | | EntroAdv | 40.4 | Not recorded | 69ddca697d0172140ec79e8a | | EAPO | 39.7 | Not recorded | 69ddca697d0172140ec79e8a | | GRPO | 39.4 | Not recorded | 69ddca697d0172140ec79e8a | | DAPO | 39.1 | Not recorded | 69ddca697d0172140ec79e8a | | Forking | 38.7 | Not recorded | 69ddca697d0172140ec79e8a | | GRPO | 38.1 | Not recorded | 69ddca697d0172140ec79e8a | | W-REINF | 37.8 | Not recorded | 69ddca697d0172140ec79e8a | | EntroReg | 37.6 | Not recorded | 69ddca697d0172140ec79e8a | | Base-1.5B | 37.5 | Not recorded | 69ddca697d0172140ec79e8a | | W-REINF | 34.2 | Not recorded | 69ddca697d0172140ec79e8a | | PRIME | 15.7 | Not recorded | 69ddca697d0172140ec79e8a | | Base-7B | 14.3 | Not recorded | 69ddca697d0172140ec79e8a | ### EAPO Avg@4: 43.9. Result ID: 69de21a847f34ce924c688ce. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Policy Entropy Method"}. This record does not assert independent reproduction. ### EntroReg Avg@4: 43.1. Result ID: 69de21a847f34ce924c688ca. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Policy Entropy Method"}. This record does not assert independent reproduction. ### Forking Avg@4: 43. Result ID: 69de21a847f34ce924c688cb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Policy Entropy Method"}. This record does not assert independent reproduction. ### DAPO Avg@4: 42.8. Result ID: 69de21a847f34ce924c688c9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Classic RLVR Method"}. This record does not assert independent reproduction. ### EntroAdv Avg@4: 42.7. Result ID: 69de21a847f34ce924c688cc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Policy Entropy Method"}. This record does not assert independent reproduction. ### KL-Cov Avg@4: 40.9. Result ID: 69de21a847f34ce924c688cd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Policy Entropy Method"}. This record does not assert independent reproduction. ### KL-Cov Avg@4: 40.5. Result ID: 69de21a847f34ce924c6889c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Policy Entropy Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### EntroAdv Avg@4: 40.4. Result ID: 69de21a847f34ce924c6889b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Policy Entropy Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### EAPO Avg@4: 39.7. Result ID: 69de21a847f34ce924c6889d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Policy Entropy Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### GRPO Avg@4: 39.4. Result ID: 69de21a847f34ce924c68897. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Method Category":"Classic RLVR Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### DAPO Avg@4: 39.1. Result ID: 69de21a847f34ce924c68898. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Classic RLVR Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### Forking Avg@4: 38.7. Result ID: 69de21a847f34ce924c6889a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Policy Entropy Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### GRPO Avg@4: 38.1. Result ID: 69de21a847f34ce924c688c8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Classic RLVR Method"}. This record does not assert independent reproduction. ### W-REINF Avg@4: 37.8. Result ID: 69de21a847f34ce924c68896. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Classic RLVR Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### EntroReg Avg@4: 37.6. Result ID: 69de21a847f34ce924c68899. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Policy Entropy Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### Base-1.5B Avg@4: 37.5. Result ID: 69de21a847f34ce924c68895. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Method Category":"Classic RLVR Method","Backbone":"DeepScaleR-1.5B-8k"}. This record does not assert independent reproduction. ### W-REINF Avg@4: 34.2. Result ID: 69de21a847f34ce924c688c7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Classic RLVR Method"}. This record does not assert independent reproduction. ### PRIME Avg@4: 15.7. Result ID: 69de21a847f34ce924c688c6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Classic RLVR Method"}. This record does not assert independent reproduction. ### Base-7B Avg@4: 14.3. Result ID: 69de21a847f34ce924c688c5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-Math-7B","Method Category":"Classic RLVR Method"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.