# Mathematical Reasoning on AIME 24 (Avg@32) Source: [Mathematical Reasoning on AIME 24 (Avg@32)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:36:31.645Z Evaluation of model performance on the AIME 2024 dataset for mathematical reasoning, specifically reporting the Avg@32 metric. 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) | AIME 24 | | Metric | Avg@32 | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 30 of 30 | | 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@32](): Higher is better; unit: SCALAR. ## Recorded results | Method | Avg@32 | Result date | Report paper | | --- | --- | --- | --- | | GSPO | 27.92 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS\* | 26.46 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS\* | 25.1 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS\* | 24.38 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS | 23.65 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS | 23.54 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS | 23.02 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO | 22.29 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS | 21.46 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS\* | 21.25 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO | 21.04 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS | 20.62 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS\* | 20.52 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO | 20.52 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS | 20.31 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS\* | 18.96 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO | 18.44 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO | 16.88 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS\* | 12.5 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS\* | 11.98 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO + CBS | 11.35 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GSPO | 10.73 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS\* | 10.52 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO | 9.58 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO + CBS | 8.96 | Not recorded | 698abc26c97a0b3e3b5de3ed | | Base Model | 8.33 | Not recorded | 698abc26c97a0b3e3b5de3ed | | GRPO + CBS | 8.12 | Not recorded | 698abc26c97a0b3e3b5de3ed | | DAPO | 6.35 | Not recorded | 698abc26c97a0b3e3b5de3ed | | Base Model | 5.1 | Not recorded | 698abc26c97a0b3e3b5de3ed | | Base Model | 1.77 | Not recorded | 698abc26c97a0b3e3b5de3ed | ### GSPO Avg@32: 27.92. Result ID: 6998d8605698665ec43c457c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS\* Avg@32: 26.46. Result ID: 6998d8605698665ec43c4578. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS\* Avg@32: 25.1. Result ID: 6998d8605698665ec43c457e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS\* Avg@32: 24.38. Result ID: 6998d8605698665ec43c457b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS Avg@32: 23.65. Result ID: 6998d8605698665ec43c4577. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS Avg@32: 23.54. Result ID: 6998d8605698665ec43c457a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS Avg@32: 23.02. Result ID: 6998d8605698665ec43c457d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### DAPO Avg@32: 22.29. Result ID: 6998d8605698665ec43c4579. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS Avg@32: 21.46. Result ID: 6998d8605698665ec43c4573. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS\* Avg@32: 21.25. Result ID: 6998d8605698665ec43c4574. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GRPO Avg@32: 21.04. Result ID: 6998d8605698665ec43c456c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS Avg@32: 20.62. Result ID: 6998d8605698665ec43c4570. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS\* Avg@32: 20.52. Result ID: 6998d8605698665ec43c456e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GRPO Avg@32: 20.52. Result ID: 6998d8605698665ec43c4576. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS Avg@32: 20.31. Result ID: 6998d8605698665ec43c456d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS\* Avg@32: 18.96. Result ID: 6998d8605698665ec43c4571. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GSPO Avg@32: 18.44. Result ID: 6998d8605698665ec43c4572. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### DAPO Avg@32: 16.88. Result ID: 6998d8605698665ec43c456f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS\* Avg@32: 12.5. Result ID: 6998d8605698665ec43c456a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS\* Avg@32: 11.98. Result ID: 6998d8605698665ec43c4564. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### GSPO + CBS Avg@32: 11.35. Result ID: 6998d8605698665ec43c4569. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### GSPO Avg@32: 10.73. Result ID: 6998d8605698665ec43c4568. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS\* Avg@32: 10.52. Result ID: 6998d8605698665ec43c4567. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### GRPO Avg@32: 9.58. Result ID: 6998d8605698665ec43c4562. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### DAPO + CBS Avg@32: 8.96. Result ID: 6998d8605698665ec43c4566. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### Base Model Avg@32: 8.33. Result ID: 6998d8605698665ec43c4575. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Base"}. This record does not assert independent reproduction. ### GRPO + CBS Avg@32: 8.12. Result ID: 6998d8605698665ec43c4563. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### DAPO Avg@32: 6.35. Result ID: 6998d8605698665ec43c4565. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. ### Base Model Avg@32: 5.1. Result ID: 6998d8605698665ec43c456b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Base"}. This record does not assert independent reproduction. ### Base Model Avg@32: 1.77. Result ID: 6998d8605698665ec43c4561. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B-Base"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.