# Mathematical Reasoning on MetaMathQA (test) Source: [Mathematical Reasoning on MetaMathQA (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:30:42.845Z Evaluation of model performance on the test split of the MetaMathQA dataset for mathematical reasoning tasks, typically measured by accuracy. 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) | MetaMathQA (test) | | Metric | Accuracy | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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 - [Accuracy](): Higher is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | DA | 53.37 | Not recorded | 698423faa8b01e76cc8ee679 | | DA | 51.78 | Not recorded | 698423faa8b01e76cc8ee679 | | RSFT | 49.73 | Not recorded | 698423faa8b01e76cc8ee679 | | DA | 49.2 | Not recorded | 698423faa8b01e76cc8ee679 | | RSFT | 48.14 | Not recorded | 698423faa8b01e76cc8ee679 | | LoRA | 47.76 | Not recorded | 68fa72551e209131cea55afa | | RSFT | 45.03 | Not recorded | 698423faa8b01e76cc8ee679 | | ResLoRAbs | 43.9 | Not recorded | 68fa72551e209131cea55afa | | PAD | 43.72 | Not recorded | 698423faa8b01e76cc8ee679 | | ResLoRAis | 43.37 | Not recorded | 68fa72551e209131cea55afa | | PAD | 42.53 | Not recorded | 698423faa8b01e76cc8ee679 | | PAD | 42.51 | Not recorded | 698423faa8b01e76cc8ee679 | | LoRA | 42.23 | Not recorded | 68fa72551e209131cea55afa | | DA | 35.43 | Not recorded | 698423faa8b01e76cc8ee679 | | PAD | 34.54 | Not recorded | 698423faa8b01e76cc8ee679 | | RSFT | 34.4 | Not recorded | 698423faa8b01e76cc8ee679 | | RSFT | 34.4 | Not recorded | 698423faa8b01e76cc8ee679 | | DA | 34.38 | Not recorded | 698423faa8b01e76cc8ee679 | | PAD | 33.27 | Not recorded | 698423faa8b01e76cc8ee679 | | ResLoRAms | 33.13 | Not recorded | 68fa72551e209131cea55afa | | PAD | 32.38 | Not recorded | 698423faa8b01e76cc8ee679 | | DA | 30.27 | Not recorded | 698423faa8b01e76cc8ee679 | | LoHA | 30.17 | Not recorded | 68fa72551e209131cea55afa | | RSFT | 28.19 | Not recorded | 698423faa8b01e76cc8ee679 | | AdaLoRA | 25.17 | Not recorded | 68fa72551e209131cea55afa | | LoKr | 21.76 | Not recorded | 68fa72551e209131cea55afa | ### DA Accuracy: 53.37. Result ID: 699605735698665ec43c82e7. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### DA Accuracy: 51.78. Result ID: 699605735698665ec43c82e1. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### RSFT Accuracy: 49.73. Result ID: 699605735698665ec43c82eb. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### DA Accuracy: 49.2. Result ID: 699605735698665ec43c82ed. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### RSFT Accuracy: 48.14. Result ID: 699605735698665ec43c82e5. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### LoRA Accuracy: 47.76. Result ID: 699210543a6c3148490c5606. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Rank":16,"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### RSFT Accuracy: 45.03. Result ID: 699605735698665ec43c82df. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### ResLoRAbs Accuracy: 43.9. Result ID: 699210543a6c3148490c560c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### PAD Accuracy: 43.72. Result ID: 699605735698665ec43c82ef. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### ResLoRAis Accuracy: 43.37. Result ID: 699210543a6c3148490c560b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### PAD Accuracy: 42.53. Result ID: 699605735698665ec43c82e9. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### PAD Accuracy: 42.51. Result ID: 699605735698665ec43c82e3. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"gsm8k"}. This record does not assert independent reproduction. ### LoRA Accuracy: 42.23. Result ID: 699210543a6c3148490c5607. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Rank":4,"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### DA Accuracy: 35.43. Result ID: 699605735698665ec43c82ee. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### PAD Accuracy: 34.54. Result ID: 699605735698665ec43c82ea. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### RSFT Accuracy: 34.4. Result ID: 699605735698665ec43c82e6. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### RSFT Accuracy: 34.4. Result ID: 699605735698665ec43c82ec. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### DA Accuracy: 34.38. Result ID: 699605735698665ec43c82e8. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-32b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### PAD Accuracy: 33.27. Result ID: 699605735698665ec43c82f0. [Reporting paper]() Recorded evaluation configuration: {"source model":"gemma3-27b-it","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### ResLoRAms Accuracy: 33.13. Result ID: 699210543a6c3148490c560d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### PAD Accuracy: 32.38. Result ID: 699605735698665ec43c82e4. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### DA Accuracy: 30.27. Result ID: 699605735698665ec43c82e2. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### LoHA Accuracy: 30.17. Result ID: 699210543a6c3148490c5609. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### RSFT Accuracy: 28.19. Result ID: 699605735698665ec43c82e0. [Reporting paper]() Recorded evaluation configuration: {"source model":"qwen3-8b","train dataset":"medmcqa"}. This record does not assert independent reproduction. ### AdaLoRA Accuracy: 25.17. Result ID: 699210543a6c3148490c5608. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. ### LoKr Accuracy: 21.76. Result ID: 699210543a6c3148490c560a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B","Fine-tuning":true}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.