# Mathematical Reasoning on Minerva-Math (Last, Best) Source: [Mathematical Reasoning on Minerva-Math (Last, Best)]() Collection index: [llms.txt]() Data updated: 2026-04-08T07:16:31.733Z Evaluation of model performance on mathematical reasoning tasks using the Minerva-Math dataset, reporting the 'Last' and 'Best' observed scores. 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 | Last Score | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [Last Score](): Higher is better; unit: SCALAR. - [Best Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | Last Score | Result date | Report paper | | --- | --- | --- | --- | | HT-MNPO | 49.63 | Not recorded | 68fa7bb81e209131cea8a23e | | HT-MNPO | 47.79 | Not recorded | 68fa7bb81e209131cea8a23e | | SPPO | 47.43 | Not recorded | 68fa7bb81e209131cea8a23e | | HT-MNPO | 47.06 | Not recorded | 68fa7bb81e209131cea8a23e | | INPO | 46.32 | Not recorded | 68fa7bb81e209131cea8a23e | | DPO | 45.96 | Not recorded | 68fa7bb81e209131cea8a23e | | TD-MNPO | 44.85 | Not recorded | 68fa7bb81e209131cea8a23e | | SFT Model | 44.12 | Not recorded | 68fa7bb81e209131cea8a23e | | SimPO | 43.38 | Not recorded | 68fa7bb81e209131cea8a23e | | REINFORCE++ | 0.1691 | Not recorded | 68fa7ba21e209131cea897d4 | | RPG-URKL | 0.1654 | Not recorded | 68fa7ba21e209131cea897d4 | | RPG-UFKL | 0.1397 | Not recorded | 68fa7ba21e209131cea897d4 | | RPG-REINFORCE-UFKL | 0.136 | Not recorded | 68fa7ba21e209131cea897d4 | | REINFORCE++-Baseline | 0.125 | Not recorded | 68fa7ba21e209131cea897d4 | | RPG-REINFORCE-URKL | 0.114 | Not recorded | 68fa7ba21e209131cea897d4 | | GRPO | 0.1029 | Not recorded | 68fa7ba21e209131cea897d4 | | DAPO | 0.0993 | Not recorded | 68fa7ba21e209131cea897d4 | ### HT-MNPO Last Score: 49.63. Result ID: 69d600c747f34ce924be22f6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Reward Model":"Skywork-Reward-V2"}. This record does not assert independent reproduction. ### HT-MNPO Last Score: 47.79. Result ID: 69d600c747f34ce924be22f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Reward Model":"ArmoRM-Llama3"}. This record does not assert independent reproduction. ### SPPO Last Score: 47.43. Result ID: 69d600c747f34ce924be22f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### HT-MNPO Last Score: 47.06. Result ID: 69d600c747f34ce924be22f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Reward Model":"Athene-RM-8B"}. This record does not assert independent reproduction. ### INPO Last Score: 46.32. Result ID: 69d600c747f34ce924be22f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DPO Last Score: 45.96. Result ID: 69d600c747f34ce924be22f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### TD-MNPO Last Score: 44.85. Result ID: 69d600c747f34ce924be22f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### SFT Model Last Score: 44.12. Result ID: 69d600c747f34ce924be22ef. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### SimPO Last Score: 43.38. Result ID: 69d600c747f34ce924be22f1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### REINFORCE++ Last Score: 0.1691. Result ID: 699af90a5698665ec48a835e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### RPG-URKL Last Score: 0.1654. Result ID: 699af90a5698665ec48a8362. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### RPG-UFKL Last Score: 0.1397. Result ID: 699af90a5698665ec48a8361. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### RPG-REINFORCE-UFKL Last Score: 0.136. Result ID: 699af90a5698665ec48a8363. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### REINFORCE++-Baseline Last Score: 0.125. Result ID: 699af90a5698665ec48a835d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### RPG-REINFORCE-URKL Last Score: 0.114. Result ID: 699af90a5698665ec48a8364. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### GRPO Last Score: 0.1029. Result ID: 699af90a5698665ec48a835f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. ### DAPO Last Score: 0.0993. Result ID: 699af90a5698665ec48a8360. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Context Length":"4k","Base Model":"Qwen-2.5-7B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.