# Mathematical Reasoning on AIME 2024 (Pass@1 Accuracy)

Source: [Mathematical Reasoning on AIME 2024 (Pass@1 Accuracy)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-2024-pass-at-1-accuracy>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-12T07:20:10.091Z

Evaluation of model performance on the American Invitational Mathematics Examination (AIME) 2024 dataset, focusing specifically on the Pass@1 accuracy 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 2024 |
| Metric | Pass@1 Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 243 of 243 |
| 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

- [Pass@1 Accuracy](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-2024-pass-at-1-accuracy.md?metric=Pass%401%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pass@1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| SeLaR | 83.33 | Not recorded | 69d884377d0172140ec77bc0 |
| SeLaR | 83.33 | Not recorded | 69d884377d0172140ec77bc0 |
| SwiR | 82.92 | Not recorded | 69d884377d0172140ec77bc0 |
| CoT | 80.42 | Not recorded | 69d884377d0172140ec77bc0 |
| CoT | 80 | Not recorded | 69d884377d0172140ec77bc0 |
| Mem2Evolve | 76.7 | Not recorded | 69ddca647d0172140ec79e05 |
| DeepSeek-R1-Distill-Qwen-32B | 76.67 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| SliderQuant | 76.67 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| CoT | 76.67 | Not recorded | 69d884377d0172140ec77bc0 |
| Qwen3-32B | 75.42 | Not recorded | 6a1e63d3dbd7150506f8543e |
| OmniOPD | 75 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Gemini-2.5-flash | 74.79 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Soft Thinking | 74.58 | Not recorded | 69d884377d0172140ec77bc0 |
| DeepSeek-R1-Distill-Qwen-14B | 73.33 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| OmniOPD | 73.13 | Not recorded | 6a1e63d3dbd7150506f8543e |
| OmniOPD | 72.08 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SFT | 71.25 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SliderQuant | 70 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| CoT | 70 | Not recorded | 69d884377d0172140ec77bc0 |
| Soft Thinking | 70 | Not recorded | 69d884377d0172140ec77bc0 |
| Alita | 70 | Not recorded | 69ddca647d0172140ec79e05 |
| OmniOPD | 68.67 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Qwen3-4B (GRPO) | 68.62 | Not recorded | 6a1e63d3dbd7150506f8543e |
| GPT-5-Chat (CoT) | 66.67 | Not recorded | 69ddca647d0172140ec79e05 |
| GPT-5-Chat (ReAct) | 66.67 | Not recorded | 69ddca647d0172140ec79e05 |
| EvoAgent | 66.67 | Not recorded | 69ddca647d0172140ec79e05 |
| AFLOW | 66.67 | Not recorded | 69ddca647d0172140ec79e05 |
| DSPy | 66.67 | Not recorded | 69ddca647d0172140ec79e05 |
| OPD | 61.45 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Qwen3-30B-A3B-Inst. | 61.04 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SFT | 60.83 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SFT | 60.83 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SwiR | 60 | Not recorded | 69d884377d0172140ec77bc0 |
| GPT-5-Chat (Direct) | 60 | Not recorded | 69ddca647d0172140ec79e05 |
| AgentVerse | 60 | Not recorded | 69ddca647d0172140ec79e05 |
| OmniQuant | 56.66 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| Meta-Reasoner | 55.5 | Not recorded | 68fa7a8b1e209131cea83d41 |
| Claude-4.5-haiku | 55.2 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SFPO | 54.17 | Not recorded | 68fa7bea1e209131cea8b458 |
| HEAL | 53.63 | Not recorded | 69b24879d030084c22e2f864 |
| SeLaR | 53.33 | Not recorded | 69d884377d0172140ec77bc0 |
| rStar-Math | 53.3 | Not recorded | 68fa7a8b1e209131cea83d41 |
| Qwen3-4B | 50.83 | Not recorded | 6a1e63d3dbd7150506f8543e |
| GRPO | 50 | Not recorded | 68fa7bea1e209131cea8b458 |
| OmniQuant | 50 | Not recorded | 69c60f2b0636e5bf4afc85dc |
| TTVS | 50 | Not recorded | 69d884407d0172140ec77c69 |
| OPD | 48.95 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Qwen3-1.7B | 48.54 | Not recorded | 6a1e63d3dbd7150506f8543e |
| OPD | 48.54 | Not recorded | 6a1e63d3dbd7150506f8543e |
| Qwen3-1.7B (GRPO) | 47.5 | Not recorded | 6a1e63d3dbd7150506f8543e |
| OmniOPD | 47.08 | Not recorded | 6a1e63d3dbd7150506f8543e |
| TTRL | 46.7 | Not recorded | 69d884407d0172140ec77c69 |
| CoT | 46.67 | Not recorded | 69d884377d0172140ec77bc0 |
| SwiR | 46.67 | Not recorded | 69d884377d0172140ec77bc0 |
| DyLAN | 46.67 | Not recorded | 69ddca647d0172140ec79e05 |
| SwarmAgentic | 46.67 | Not recorded | 69ddca647d0172140ec79e05 |
| rStar-Math | 43.3 | Not recorded | 68fa7a8b1e209131cea83d41 |
| MCTS-RAP | 40.2 | Not recorded | 68fa7a8b1e209131cea83d41 |
| CoT | 40 | Not recorded | 69d884377d0172140ec77bc0 |
| AutoAgents | 40 | Not recorded | 69ddca647d0172140ec79e05 |
| Full Fine-Tuning | 37.03 | Not recorded | 6a2b92b6d33dfb23e783b13e |
| TTVS | 36.7 | Not recorded | 69d884407d0172140ec77c69 |
| Soft Thinking | 36.67 | Not recorded | 69d884377d0172140ec77bc0 |
| NFPO | 36.4 | Not recorded | 6a0e9177f984739ed67389fa |
| Curriculum SFT | 36.27 | Not recorded | 69b24879d030084c22e2f864 |
| SFT | 35.42 | Not recorded | 6a1e63d3dbd7150506f8543e |
| SFPO | 35 | Not recorded | 68fa7bea1e209131cea8b458 |
| DPPO | 34.7 | Not recorded | 6a0e9177f984739ed67389fa |
| SFT | 34.58 | Not recorded | 6a1e63d3dbd7150506f8543e |
| LoRA-α | 34.21 | Not recorded | 6a2b92b6d33dfb23e783b13e |
| GRPO | 34.2 | Not recorded | 68fa7bea1e209131cea8b458 |
| LoRA (10× LR) | 33.69 | Not recorded | 6a2b92b6d33dfb23e783b13e |
| Meta-Reasoner | 33.3 | Not recorded | 68fa7a8b1e209131cea83d41 |
| SFT | 33.23 | Not recorded | 69b24879d030084c22e2f864 |
| SFPO | 32.5 | Not recorded | 68fa7bea1e209131cea8b458 |
| HEAL | 30.1 | Not recorded | 69b24879d030084c22e2f864 |
| GRPO | 30 | Not recorded | 68fa7bea1e209131cea8b458 |
| LoRA-α | 28.69 | Not recorded | 6a2b92b6d33dfb23e783b13e |
| Base | 28.33 | Not recorded | 68fa7bea1e209131cea8b458 |
| LoRA (10× LR) | 28.29 | Not recorded | 6a2b92b6d33dfb23e783b13e |

### SeLaR

Pass@1 Accuracy: 83.33. Result ID: 69d8c24647f34ce924c1f6ab.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### SeLaR

Pass@1 Accuracy: 83.33. Result ID: 69d8c24647f34ce924c1f6b0.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-32B"}. This record does not assert independent reproduction.

### SwiR

Pass@1 Accuracy: 82.92. Result ID: 69d8c24647f34ce924c1f6af.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be21e209131cea8b095.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-32B"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 80.42. Result ID: 69d8c24647f34ce924c1f6ac.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-32B","Decoding Strategy":"Sampling"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 80. Result ID: 69d8c24647f34ce924c1f6ad.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-32B","Decoding Strategy":"Greedy"}. This record does not assert independent reproduction.

### Mem2Evolve

Pass@1 Accuracy: 76.7. Result ID: 69de1d6647f34ce924c6653b.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### DeepSeek-R1-Distill-Qwen-32B

Pass@1 Accuracy: 76.67. Result ID: 69c6a24047f34ce924b10960.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a91e209131cea6e3eb.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-32B","#Bits":"W16A16"}. This record does not assert independent reproduction.

### SliderQuant

Pass@1 Accuracy: 76.67. Result ID: 69c6a24047f34ce924b10964.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-32B","#Bits":"W4A16"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 76.67. Result ID: 69d8c24647f34ce924c1f6a7.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Decoding Strategy":"Sampling"}. This record does not assert independent reproduction.

### Qwen3-32B

Pass@1 Accuracy: 75.42. Result ID: 6a1eb8a3fbc373592c19cbc5.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Student Model":"-","Teacher Model":"Qwen3-32B","Method":"Base"}. This record does not assert independent reproduction.

### OmniOPD

Pass@1 Accuracy: 75. Result ID: 6a1eb8a3fbc373592c19cbd5.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-30B-A3B-Inst.","Method":"OmniOPD"}. This record does not assert independent reproduction.

### Gemini-2.5-flash

Pass@1 Accuracy: 74.79. Result ID: 6a1eb8a3fbc373592c19cbc8.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a642.md>)

Recorded evaluation configuration: {"Student Model":"-","Teacher Model":"Gemini-2.5-flash","Method":"Base"}. This record does not assert independent reproduction.

### Soft Thinking

Pass@1 Accuracy: 74.58. Result ID: 69d8c24647f34ce924c1f6ae.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78ba1e209131cea7a14a.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-32B"}. This record does not assert independent reproduction.

### DeepSeek-R1-Distill-Qwen-14B

Pass@1 Accuracy: 73.33. Result ID: 69c6a24047f34ce924b1095b.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a91e209131cea6e3eb.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-14B","#Bits":"W16A16"}. This record does not assert independent reproduction.

### OmniOPD

Pass@1 Accuracy: 73.13. Result ID: 6a1eb8a3fbc373592c19cbd9.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Gemini-2.5-flash","Method":"OmniOPD"}. This record does not assert independent reproduction.

### OmniOPD

Pass@1 Accuracy: 72.08. Result ID: 6a1eb8a3fbc373592c19cbd7.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Claude-4.5-haiku","Method":"OmniOPD"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 71.25. Result ID: 6a1eb8a3fbc373592c19cbd8.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bb1e209131ce9faed1.md>)

[Implementation](<https://github.com/harvardnlp/seq2seq-attn>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Gemini-2.5-flash","Method":"SFT"}. This record does not assert independent reproduction.

### SliderQuant

Pass@1 Accuracy: 70. Result ID: 69c6a24047f34ce924b1095f.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-14B","#Bits":"W4A16"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 70. Result ID: 69d8c24647f34ce924c1f6a8.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Decoding Strategy":"Greedy"}. This record does not assert independent reproduction.

### Soft Thinking

Pass@1 Accuracy: 70. Result ID: 69d8c24647f34ce924c1f6a9.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78ba1e209131cea7a14a.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### Alita

Pass@1 Accuracy: 70. Result ID: 69de1d6647f34ce924c66537.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78da1e209131cea7acec.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### OmniOPD

Pass@1 Accuracy: 68.67. Result ID: 6a1eb8a3fbc373592c19cbd2.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-32B","Method":"OmniOPD"}. This record does not assert independent reproduction.

### Qwen3-4B (GRPO)

Pass@1 Accuracy: 68.62. Result ID: 6a1eb8a3fbc373592c19cbcf.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"-","Method":"GRPO"}. This record does not assert independent reproduction.

### GPT-5-Chat (CoT)

Pass@1 Accuracy: 66.67. Result ID: 69de1d6647f34ce924c66531.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### GPT-5-Chat (ReAct)

Pass@1 Accuracy: 66.67. Result ID: 69de1d6647f34ce924c66532.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### EvoAgent

Pass@1 Accuracy: 66.67. Result ID: 69de1d6647f34ce924c66534.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa775b1e209131cea72ad7.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### AFLOW

Pass@1 Accuracy: 66.67. Result ID: 69de1d6647f34ce924c66535.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78111e209131cea769ef.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### DSPy

Pass@1 Accuracy: 66.67. Result ID: 69de1d6647f34ce924c66536.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d01e209131cea4d60e.md>)

[Implementation](<https://github.com/stanfordnlp/dspy>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### OPD

Pass@1 Accuracy: 61.45. Result ID: 6a1eb8a3fbc373592c19cbd1.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-32B","Method":"OPD"}. This record does not assert independent reproduction.

### Qwen3-30B-A3B-Inst.

Pass@1 Accuracy: 61.04. Result ID: 6a1eb8a3fbc373592c19cbc6.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Student Model":"-","Teacher Model":"Qwen3-30B-A3B-Inst.","Method":"Base"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 60.83. Result ID: 6a1eb8a3fbc373592c19cbd0.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bb1e209131ce9faed1.md>)

[Implementation](<https://github.com/harvardnlp/seq2seq-attn>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-32B","Method":"SFT"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 60.83. Result ID: 6a1eb8a3fbc373592c19cbd6.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bb1e209131ce9faed1.md>)

[Implementation](<https://github.com/harvardnlp/seq2seq-attn>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Claude-4.5-haiku","Method":"SFT"}. This record does not assert independent reproduction.

### SwiR

Pass@1 Accuracy: 60. Result ID: 69d8c24647f34ce924c1f6aa.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be21e209131cea8b095.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### GPT-5-Chat (Direct)

Pass@1 Accuracy: 60. Result ID: 69de1d6647f34ce924c66530.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### AgentVerse

Pass@1 Accuracy: 60. Result ID: 69de1d6647f34ce924c66538.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f31e209131cea4e4c3.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### OmniQuant

Pass@1 Accuracy: 56.66. Result ID: 69c6a24047f34ce924b10963.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72941e209131cea5741d.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-32B","#Bits":"W4A16"}. This record does not assert independent reproduction.

### Meta-Reasoner

Pass@1 Accuracy: 55.5. Result ID: 69fdbf4c47f34ce924e17520.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### Claude-4.5-haiku

Pass@1 Accuracy: 55.2. Result ID: 6a1eb8a3fbc373592c19cbc7.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"-","Teacher Model":"Claude-4.5-haiku","Method":"Base"}. This record does not assert independent reproduction.

### SFPO

Pass@1 Accuracy: 54.17. Result ID: 69b91ea347f34ce924a20e72.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

Recorded evaluation configuration: {"Model":"DS-distilled-Qwen-7B"}. This record does not assert independent reproduction.

### HEAL

Pass@1 Accuracy: 53.63. Result ID: 69b2707e6dbb7075b95ccb35.

[Reporting paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

Recorded evaluation configuration: {"Models":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### SeLaR

Pass@1 Accuracy: 53.33. Result ID: 69d8c24647f34ce924c1f6a6.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### rStar-Math

Pass@1 Accuracy: 53.3. Result ID: 69fdbf4c47f34ce924e17519.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa768e1e209131cea6db2e.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-Math-7B"}. This record does not assert independent reproduction.

### Qwen3-4B

Pass@1 Accuracy: 50.83. Result ID: 6a1eb8a3fbc373592c19cbce.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"-","Method":"Base"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Accuracy: 50. Result ID: 69b91ea347f34ce924a20e71.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Model":"DS-distilled-Qwen-7B"}. This record does not assert independent reproduction.

### OmniQuant

Pass@1 Accuracy: 50. Result ID: 69c6a24047f34ce924b1095e.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f2b0636e5bf4afc85dc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72941e209131cea5741d.md>)

Recorded evaluation configuration: {"Model":"DeepSeek-R1-Distill-Qwen-14B","#Bits":"W4A16"}. This record does not assert independent reproduction.

### TTVS

Pass@1 Accuracy: 50. Result ID: 69d8c59747f34ce924c210df.

[Reporting paper](<https://www.sota2.com/research/paper/69d884407d0172140ec77c69.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d884407d0172140ec77c69.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### OPD

Pass@1 Accuracy: 48.95. Result ID: 6a1eb8a3fbc373592c19cbcc.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-1.7B","Teacher Model":"Qwen3-32B","Method":"OPD"}. This record does not assert independent reproduction.

### Qwen3-1.7B

Pass@1 Accuracy: 48.54. Result ID: 6a1eb8a3fbc373592c19cbc9.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-1.7B","Teacher Model":"-","Method":"Base"}. This record does not assert independent reproduction.

### OPD

Pass@1 Accuracy: 48.54. Result ID: 6a1eb8a3fbc373592c19cbd4.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-30B-A3B-Inst.","Method":"OPD"}. This record does not assert independent reproduction.

### Qwen3-1.7B (GRPO)

Pass@1 Accuracy: 47.5. Result ID: 6a1eb8a3fbc373592c19cbca.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Student Model":"Qwen3-1.7B","Teacher Model":"-","Method":"GRPO"}. This record does not assert independent reproduction.

### OmniOPD

Pass@1 Accuracy: 47.08. Result ID: 6a1eb8a3fbc373592c19cbcd.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

Recorded evaluation configuration: {"Student Model":"Qwen3-1.7B","Teacher Model":"Qwen3-32B","Method":"OmniOPD"}. This record does not assert independent reproduction.

### TTRL

Pass@1 Accuracy: 46.7. Result ID: 69d8c59747f34ce924c210de.

[Reporting paper](<https://www.sota2.com/research/paper/69d884407d0172140ec77c69.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79ce1e209131cea7f7aa.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 46.67. Result ID: 69d8c24647f34ce924c1f6a2.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B","Decoding Strategy":"Sampling"}. This record does not assert independent reproduction.

### SwiR

Pass@1 Accuracy: 46.67. Result ID: 69d8c24647f34ce924c1f6a5.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be21e209131cea8b095.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### DyLAN

Pass@1 Accuracy: 46.67. Result ID: 69de1d6647f34ce924c66533.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75b01e209131cea69532.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### SwarmAgentic

Pass@1 Accuracy: 46.67. Result ID: 69de1d6647f34ce924c6653a.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79801e209131cea7dda7.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### rStar-Math

Pass@1 Accuracy: 43.3. Result ID: 69fdbf4c47f34ce924e1751a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa768e1e209131cea6db2e.md>)

Recorded evaluation configuration: {"Base Model":"Phi3-mini-3.8B"}. This record does not assert independent reproduction.

### MCTS-RAP

Pass@1 Accuracy: 40.2. Result ID: 69fdbf4c47f34ce924e1751d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e480.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen-7B"}. This record does not assert independent reproduction.

### CoT

Pass@1 Accuracy: 40. Result ID: 69d8c24647f34ce924c1f6a3.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e571e209131cea3f27b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B","Decoding Strategy":"Greedy"}. This record does not assert independent reproduction.

### AutoAgents

Pass@1 Accuracy: 40. Result ID: 69de1d6647f34ce924c66539.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca647d0172140ec79e05.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73081e209131cea59f36.md>)

Recorded evaluation configuration: {"Backbone":"GPT-5-Chat"}. This record does not assert independent reproduction.

### Full Fine-Tuning

Pass@1 Accuracy: 37.03. Result ID: 6a2bb31ffbc373592c29ac13.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

Recorded evaluation configuration: {"Rank (r)":"N/A","n":"64"}. This record does not assert independent reproduction.

### TTVS

Pass@1 Accuracy: 36.7. Result ID: 69d8c59747f34ce924c210dc.

[Reporting paper](<https://www.sota2.com/research/paper/69d884407d0172140ec77c69.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d884407d0172140ec77c69.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-4B"}. This record does not assert independent reproduction.

### Soft Thinking

Pass@1 Accuracy: 36.67. Result ID: 69d8c24647f34ce924c1f6a4.

[Reporting paper](<https://www.sota2.com/research/paper/69d884377d0172140ec77bc0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78ba1e209131cea7a14a.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### NFPO

Pass@1 Accuracy: 36.4. Result ID: 6a0ef24afbc373592c074e32.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9177f984739ed67389fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0e9177f984739ed67389fa.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-8B-Base","Algorithm":"NFPO"}. This record does not assert independent reproduction.

### Curriculum SFT

Pass@1 Accuracy: 36.27. Result ID: 69b2707e6dbb7075b95ccb34.

[Reporting paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78d41e209131cea7aa25.md>)

Recorded evaluation configuration: {"Models":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 35.42. Result ID: 6a1eb8a3fbc373592c19cbcb.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bb1e209131ce9faed1.md>)

[Implementation](<https://github.com/harvardnlp/seq2seq-attn>)

Recorded evaluation configuration: {"Student Model":"Qwen3-1.7B","Teacher Model":"Qwen3-32B","Method":"SFT"}. This record does not assert independent reproduction.

### SFPO

Pass@1 Accuracy: 35. Result ID: 69b91ea347f34ce924a20e6c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B"}. This record does not assert independent reproduction.

### DPPO

Pass@1 Accuracy: 34.7. Result ID: 6a0ef24afbc373592c074e31.

[Reporting paper](<https://www.sota2.com/research/paper/6a0e9177f984739ed67389fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984247ba8b01e76cc8eea0f.md>)

Recorded evaluation configuration: {"Base Model":"Qwen3-8B-Base","Algorithm":"DPPO"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 34.58. Result ID: 6a1eb8a3fbc373592c19cbd3.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63d3dbd7150506f8543e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bb1e209131ce9faed1.md>)

[Implementation](<https://github.com/harvardnlp/seq2seq-attn>)

Recorded evaluation configuration: {"Student Model":"Qwen3-4B","Teacher Model":"Qwen3-30B-A3B-Inst.","Method":"SFT"}. This record does not assert independent reproduction.

### LoRA-α

Pass@1 Accuracy: 34.21. Result ID: 6a2bb31ffbc373592c29ac19.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

Recorded evaluation configuration: {"Rank (r)":"256","n":"64"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Accuracy: 34.2. Result ID: 69b91ea347f34ce924a20e6b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Model":"Qwen2.5-Math-7B"}. This record does not assert independent reproduction.

### LoRA (10× LR)

Pass@1 Accuracy: 33.69. Result ID: 6a2bb31ffbc373592c29ac18.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

Recorded evaluation configuration: {"Rank (r)":"256","n":"64"}. This record does not assert independent reproduction.

### Meta-Reasoner

Pass@1 Accuracy: 33.3. Result ID: 69fdbf4c47f34ce924e1751f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a8b1e209131cea83d41.md>)

Recorded evaluation configuration: {"Base Model":"GPT-4o"}. This record does not assert independent reproduction.

### SFT

Pass@1 Accuracy: 33.23. Result ID: 69b2707e6dbb7075b95ccb33.

[Reporting paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f701e209131cea45952.md>)

Recorded evaluation configuration: {"Models":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### SFPO

Pass@1 Accuracy: 32.5. Result ID: 69b91ea347f34ce924a20e6f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

Recorded evaluation configuration: {"Model":"DS-distilled-Qwen-1.5B"}. This record does not assert independent reproduction.

### HEAL

Pass@1 Accuracy: 30.1. Result ID: 69b2707e6dbb7075b95ccb3a.

[Reporting paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b24879d030084c22e2f864.md>)

Recorded evaluation configuration: {"Models":"Qwen3-4B-Base"}. This record does not assert independent reproduction.

### GRPO

Pass@1 Accuracy: 30. Result ID: 69b91ea347f34ce924a20e6e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a395.md>)

[Implementation](<https://github.com/deepseek-ai/DeepSeek-Math>)

Recorded evaluation configuration: {"Model":"DS-distilled-Qwen-1.5B"}. This record does not assert independent reproduction.

### LoRA-α

Pass@1 Accuracy: 28.69. Result ID: 6a2bb31ffbc373592c29ac16.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

Recorded evaluation configuration: {"Rank (r)":"64","n":"64"}. This record does not assert independent reproduction.

### Base

Pass@1 Accuracy: 28.33. Result ID: 69b91ea347f34ce924a20e70.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bea1e209131cea8b458.md>)

Recorded evaluation configuration: {"Model":"DS-distilled-Qwen-7B"}. This record does not assert independent reproduction.

### LoRA (10× LR)

Pass@1 Accuracy: 28.29. Result ID: 6a2bb31ffbc373592c29ac15.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b92b6d33dfb23e783b13e.md>)

Recorded evaluation configuration: {"Rank (r)":"64","n":"64"}. This record does not assert independent reproduction.

Page 1 of 4. This is an index page, not the full collection.

- [Next index page](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-2024-pass-at-1-accuracy.md?metric=Pass%401%20Accuracy&page=2>): Follow this link to discover more pages.
