# Mathematical Reasoning on AMC 2023 (Accuracy (avg@32), Average) Source: [Mathematical Reasoning on AMC 2023 (Accuracy (avg@32), Average)]() Collection index: [llms.txt]() Data updated: 2026-06-17T06:32:22.910Z Evaluation of model performance on the AMC 2023 dataset for Mathematical Reasoning, specifically measured using Accuracy (avg@32) and Average metrics. 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) | AMC 2023 | | Metric | Accuracy (avg@32) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 36 of 36 | | 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 (avg@32)](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (avg@32) | Result date | Report paper | | --- | --- | --- | --- | | SRaR | 94.77 | Not recorded | 6a0bee98dbc836b87d82c42c | | RaR | 94.22 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-7B + IC-DAPO | 93.7 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DAPO | 93.67 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-7B + CE-GPPO | 93.3 | Not recorded | 69b0f7132d603e6de3e45dc4 | | SRaR | 92.73 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-7B + DAPO | 92.3 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DS-R1-Distill-Qwen-7B | 90.6 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DAPO | 90.23 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-7B + GRPO | 88.8 | Not recorded | 69b0f7132d603e6de3e45dc4 | | RaR | 88.2 | Not recorded | 6a0bee98dbc836b87d82c42c | | RGR-GRPO | 87.42 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-1.5B + IC-DAPO | 86.2 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DS-R1-Distill-Qwen-1.5B + GSPO | 85.9 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DS-R1-Distill-Qwen-1.5B + CE-GPPO | 85.8 | Not recorded | 69b0f7132d603e6de3e45dc4 | | DS-R1-Distill-Qwen-1.5B + DAPO | 84.4 | Not recorded | 69b0f7132d603e6de3e45dc4 | | GRPO-VPS | 83.2 | Not recorded | 6a0bee98dbc836b87d82c42c | | RGR-GRPO | 81.09 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-1.5B + CISPO | 80.9 | Not recorded | 69b0f7132d603e6de3e45dc4 | | GRPO | 80.78 | Not recorded | 6a0bee98dbc836b87d82c42c | | GRPO-VPS | 80.08 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-1.5B + GRPO | 79.3 | Not recorded | 69b0f7132d603e6de3e45dc4 | | GRPO | 79.06 | Not recorded | 6a0bee98dbc836b87d82c42c | | DS-R1-Distill-Qwen-1.5B | 73.7 | Not recorded | 69b0f7132d603e6de3e45dc4 | | Baseline | 72.03 | Not recorded | 6a0bee98dbc836b87d82c42c | | Baseline | 71.48 | Not recorded | 6a0bee98dbc836b87d82c42c | | GRPO | 68.91 | Not recorded | 6a322a2ad33dfb23e783eb93 | | RF++ w/ NTF | 68.91 | Not recorded | 6a322a2ad33dfb23e783eb93 | | GRPO | 64.84 | Not recorded | 6a322a2ad33dfb23e783eb93 | | C-RF w/ NTF | 61.95 | Not recorded | 6a322a2ad33dfb23e783eb93 | | RF++ | 58.36 | Not recorded | 6a322a2ad33dfb23e783eb93 | | GRPO | 57.11 | Not recorded | 6a322a2ad33dfb23e783eb93 | | RF++ w/ NTF | 55.47 | Not recorded | 6a322a2ad33dfb23e783eb93 | | C-RF w/ NTF | 55.47 | Not recorded | 6a322a2ad33dfb23e783eb93 | | RF++ | 54.77 | Not recorded | 6a322a2ad33dfb23e783eb93 | | GRPO | 52.19 | Not recorded | 6a322a2ad33dfb23e783eb93 | ### SRaR Accuracy (avg@32): 94.77. Result ID: 6a0c30effbc373592c02b125. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### RaR Accuracy (avg@32): 94.22. Result ID: 6a0c30effbc373592c02b124. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-7B + IC-DAPO Accuracy (avg@32): 93.7. Result ID: 69b135a56dbb7075b95bf077. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-7B","Zero-shot mode":"true","Time/Step (s)":"315.6"}. This record does not assert independent reproduction. ### DAPO Accuracy (avg@32): 93.67. Result ID: 6a0c30effbc373592c02b123. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-7B + CE-GPPO Accuracy (avg@32): 93.3. Result ID: 69b135a56dbb7075b95bf076. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-7B","Zero-shot mode":"true","Time/Step (s)":"292.5"}. This record does not assert independent reproduction. ### SRaR Accuracy (avg@32): 92.73. Result ID: 6a0c30effbc373592c02b11e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-7B + DAPO Accuracy (avg@32): 92.3. Result ID: 69b135a56dbb7075b95bf075. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-7B","Zero-shot mode":"true","Time/Step (s)":"303.1"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-7B Accuracy (avg@32): 90.6. Result ID: 69b135a56dbb7075b95bf073. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen-7B","Zero-shot mode":"true"}. This record does not assert independent reproduction. ### DAPO Accuracy (avg@32): 90.23. Result ID: 6a0c30effbc373592c02b11c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-7B + GRPO Accuracy (avg@32): 88.8. Result ID: 69b135a56dbb7075b95bf074. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen-7B","Zero-shot mode":"true","Time/Step (s)":"305.6"}. This record does not assert independent reproduction. ### RaR Accuracy (avg@32): 88.2. Result ID: 6a0c30effbc373592c02b11d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### RGR-GRPO Accuracy (avg@32): 87.42. Result ID: 6a0c30effbc373592c02b122. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + IC-DAPO Accuracy (avg@32): 86.2. Result ID: 69b135a56dbb7075b95bf072. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"477.2"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + GSPO Accuracy (avg@32): 85.9. Result ID: 69b135a56dbb7075b95bf070. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"437.3"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + CE-GPPO Accuracy (avg@32): 85.8. Result ID: 69b135a56dbb7075b95bf071. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"464.0"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + DAPO Accuracy (avg@32): 84.4. Result ID: 69b135a56dbb7075b95bf06e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"459.6"}. This record does not assert independent reproduction. ### GRPO-VPS Accuracy (avg@32): 83.2. Result ID: 6a0c30effbc373592c02b11a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### RGR-GRPO Accuracy (avg@32): 81.09. Result ID: 6a0c30effbc373592c02b11b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + CISPO Accuracy (avg@32): 80.9. Result ID: 69b135a56dbb7075b95bf06f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"466.3"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 80.78. Result ID: 6a0c30effbc373592c02b120. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### GRPO-VPS Accuracy (avg@32): 80.08. Result ID: 6a0c30effbc373592c02b121. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B + GRPO Accuracy (avg@32): 79.3. Result ID: 69b135a56dbb7075b95bf06d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true","Time/Step (s)":"457.4"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 79.06. Result ID: 6a0c30effbc373592c02b119. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### DS-R1-Distill-Qwen-1.5B Accuracy (avg@32): 73.7. Result ID: 69b135a56dbb7075b95bf06c. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen-1.5B","Zero-shot mode":"true"}. This record does not assert independent reproduction. ### Baseline Accuracy (avg@32): 72.03. Result ID: 6a0c30effbc373592c02b118. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B"}. This record does not assert independent reproduction. ### Baseline Accuracy (avg@32): 71.48. Result ID: 6a0c30effbc373592c02b11f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 68.91. Result ID: 6a323f5efbc373592c2f0f49. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B","G":"16"}. This record does not assert independent reproduction. ### RF++ w/ NTF Accuracy (avg@32): 68.91. Result ID: 6a323f5efbc373592c2f0f4c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B","G":"1"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 64.84. Result ID: 6a323f5efbc373592c2f0f4a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B","G":"2"}. This record does not assert independent reproduction. ### C-RF w/ NTF Accuracy (avg@32): 61.95. Result ID: 6a323f5efbc373592c2f0f4d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B","G":"1"}. This record does not assert independent reproduction. ### RF++ Accuracy (avg@32): 58.36. Result ID: 6a323f5efbc373592c2f0f4b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B","G":"8"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 57.11. Result ID: 6a323f5efbc373592c2f0f44. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-1.5B","G":"16"}. This record does not assert independent reproduction. ### RF++ w/ NTF Accuracy (avg@32): 55.47. Result ID: 6a323f5efbc373592c2f0f47. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-1.5B","G":"1"}. This record does not assert independent reproduction. ### C-RF w/ NTF Accuracy (avg@32): 55.47. Result ID: 6a323f5efbc373592c2f0f48. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-1.5B","G":"1"}. This record does not assert independent reproduction. ### RF++ Accuracy (avg@32): 54.77. Result ID: 6a323f5efbc373592c2f0f46. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-1.5B","G":"8"}. This record does not assert independent reproduction. ### GRPO Accuracy (avg@32): 52.19. Result ID: 6a323f5efbc373592c2f0f45. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-Math-1.5B","G":"2"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.