# Mathematical Reasoning on MATH 500 (Accuracy, Memory Usage) Source: [Mathematical Reasoning on MATH 500 (Accuracy, Memory Usage)]() Collection index: [llms.txt]() Data updated: 2026-05-22T08:18:23.597Z Evaluation of model performance on mathematical reasoning problems using the MATH 500 dataset, reporting standard accuracy and peak memory consumption. 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) | MATH 500 | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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: PERCENT. - [Peak Memory (GiB)](): Lower is better; unit: SIZE. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | FullKV | 81.2 | Not recorded | 6a0fe32cf984739ed6739f06 | | FullKV | 79.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | ArborKV | 77.6 | Not recorded | 6a0fe32cf984739ed6739f06 | | StreamingLLM | 76.4 | Not recorded | 6a0fe32cf984739ed6739f06 | | ThinKV | 76 | Not recorded | 6a0fe32cf984739ed6739f06 | | ArborKV | 75.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | H2O | 74.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | StreamingLLM | 74.2 | Not recorded | 6a0fe32cf984739ed6739f06 | | ThinKV | 73.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | H2O | 72.2 | Not recorded | 6a0fe32cf984739ed6739f06 | | ArborKV | 67.4 | Not recorded | 6a0fe32cf984739ed6739f06 | | ThinKV | 65.4 | Not recorded | 6a0fe32cf984739ed6739f06 | | ArborKV | 64.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | StreamingLLM | 62.8 | Not recorded | 6a0fe32cf984739ed6739f06 | | StreamingLLM | 62.4 | Not recorded | 6a0fe32cf984739ed6739f06 | | ThinKV | 61.6 | Not recorded | 6a0fe32cf984739ed6739f06 | | H2O | 61.6 | Not recorded | 6a0fe32cf984739ed6739f06 | | H2O | 60.2 | Not recorded | 6a0fe32cf984739ed6739f06 | ### FullKV Accuracy: 81.2. Result ID: 6a10114bfbc373592c08f644. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"1.00"}. This record does not assert independent reproduction. ### FullKV Accuracy: 79.8. Result ID: 6a10114bfbc373592c08f63b. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"1.00"}. This record does not assert independent reproduction. ### ArborKV Accuracy: 77.6. Result ID: 6a10114bfbc373592c08f648. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.50"}. This record does not assert independent reproduction. ### StreamingLLM Accuracy: 76.4. Result ID: 6a10114bfbc373592c08f646. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.50"}. This record does not assert independent reproduction. ### ThinKV Accuracy: 76. Result ID: 6a10114bfbc373592c08f647. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.50"}. This record does not assert independent reproduction. ### ArborKV Accuracy: 75.8. Result ID: 6a10114bfbc373592c08f63f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.50"}. This record does not assert independent reproduction. ### H2O Accuracy: 74.8. Result ID: 6a10114bfbc373592c08f645. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.50"}. This record does not assert independent reproduction. ### StreamingLLM Accuracy: 74.2. Result ID: 6a10114bfbc373592c08f63d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.50"}. This record does not assert independent reproduction. ### ThinKV Accuracy: 73.8. Result ID: 6a10114bfbc373592c08f63e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.50"}. This record does not assert independent reproduction. ### H2O Accuracy: 72.2. Result ID: 6a10114bfbc373592c08f63c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.50"}. This record does not assert independent reproduction. ### ArborKV Accuracy: 67.4. Result ID: 6a10114bfbc373592c08f64c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.25"}. This record does not assert independent reproduction. ### ThinKV Accuracy: 65.4. Result ID: 6a10114bfbc373592c08f64b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.25"}. This record does not assert independent reproduction. ### ArborKV Accuracy: 64.8. Result ID: 6a10114bfbc373592c08f643. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.25"}. This record does not assert independent reproduction. ### StreamingLLM Accuracy: 62.8. Result ID: 6a10114bfbc373592c08f64a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.25"}. This record does not assert independent reproduction. ### StreamingLLM Accuracy: 62.4. Result ID: 6a10114bfbc373592c08f641. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.25"}. This record does not assert independent reproduction. ### ThinKV Accuracy: 61.6. Result ID: 6a10114bfbc373592c08f642. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.25"}. This record does not assert independent reproduction. ### H2O Accuracy: 61.6. Result ID: 6a10114bfbc373592c08f649. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen-2.5-7B","rho":"0.25"}. This record does not assert independent reproduction. ### H2O Accuracy: 60.2. Result ID: 6a10114bfbc373592c08f640. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B","rho":"0.25"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.