# Mathematical Reasoning on Aggregate (AMC23, AIME24, MATH-500, Minerva, Olympiad)

Source: [Mathematical Reasoning on Aggregate (AMC23, AIME24, MATH-500, Minerva, Olympiad)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aggregate-amc23-aime24-math-500-minerva-olympiad>)

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

Data updated: 2026-05-11T09:40:22.578Z

Evaluation of Mathematical Reasoning performance across an aggregate benchmark comprising AMC23, AIME24, MATH-500, Minerva, and Olympiad datasets, reporting accuracy deltas, token deltas, and Intelligence Per Token (IPT).

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) | Aggregate AMC23, AIME24, MATH-500, Minerva, Olympiad |
| Metric | Intelligence Per Token (IPT) |
| 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

- [Intelligence Per Token (IPT)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aggregate-amc23-aime24-math-500-minerva-olympiad.md?metric=Intelligence%20Per%20Token%20(IPT)>): Higher is better; unit: SCALAR.

- [Accuracy Delta (%)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aggregate-amc23-aime24-math-500-minerva-olympiad.md?metric=Accuracy%20Delta%20(%25)>): Higher is better; unit: PERCENT.

- [Token Delta (%)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aggregate-amc23-aime24-math-500-minerva-olympiad.md?metric=Token%20Delta%20(%25)>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Intelligence Per Token (IPT) | Result date | Report paper |
| --- | --- | --- | --- |
| TRiMS | 68.2 | Not recorded | 69bb83200636e5bf4afc1b3e |
| ExpThink | 51.04 | Not recorded | 6a0162afdbc836b87d823244 |
| TRiMS | 45.1 | Not recorded | 69bb83200636e5bf4afc1b3e |
| ExpThink | 32.84 | Not recorded | 6a0162afdbc836b87d823244 |
| LC-R1 | 28.4 | Not recorded | 69bb83200636e5bf4afc1b3e |
| Laser | 27.8 | Not recorded | 69bb83200636e5bf4afc1b3e |
| LC-R1 | 23.47 | Not recorded | 6a0162afdbc836b87d823244 |
| Laser | 23.3 | Not recorded | 6a0162afdbc836b87d823244 |
| ExpThink | 23.29 | Not recorded | 6a0162afdbc836b87d823244 |
| JET | 21.1 | Not recorded | 69bb83200636e5bf4afc1b3e |
| AdaptThink | 19.52 | Not recorded | 6a0162afdbc836b87d823244 |
| AdaptThink | 19.3 | Not recorded | 69bb83200636e5bf4afc1b3e |
| AutoThink | 18.63 | Not recorded | 6a0162afdbc836b87d823244 |
| AdaptThink | 18.5 | Not recorded | 69bb83200636e5bf4afc1b3e |
| AdaptThink | 18.04 | Not recorded | 6a0162afdbc836b87d823244 |
| AutoThink | 17.9 | Not recorded | 69bb83200636e5bf4afc1b3e |
| JET | 16.7 | Not recorded | 69bb83200636e5bf4afc1b3e |
| Laser | 15.8 | Not recorded | 69bb83200636e5bf4afc1b3e |
| AutoThink | 14.9 | Not recorded | 69bb83200636e5bf4afc1b3e |
| LC-R1 | 14.49 | Not recorded | 6a0162afdbc836b87d823244 |
| JET | 14.35 | Not recorded | 6a0162afdbc836b87d823244 |
| LC-R1 | 13.6 | Not recorded | 69bb83200636e5bf4afc1b3e |
| Laser | 13.35 | Not recorded | 6a0162afdbc836b87d823244 |
| AutoThink | 13.25 | Not recorded | 6a0162afdbc836b87d823244 |
| Vanilla | 11.2 | Not recorded | 69bb83200636e5bf4afc1b3e |
| Vanilla | 11.15 | Not recorded | 6a0162afdbc836b87d823244 |
| JET | 10.86 | Not recorded | 6a0162afdbc836b87d823244 |
| Vanilla | 8.64 | Not recorded | 6a0162afdbc836b87d823244 |
| Vanilla | 7.23 | Not recorded | 6a0162afdbc836b87d823244 |
| Vanilla | 7.2 | Not recorded | 69bb83200636e5bf4afc1b3e |

### TRiMS

Intelligence Per Token (IPT): 68.2. Result ID: 69bbbf8c47f34ce924a5e538.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

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

### ExpThink

Intelligence Per Token (IPT): 51.04. Result ID: 6a01a40447f34ce924e3e774.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

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

### TRiMS

Intelligence Per Token (IPT): 45.1. Result ID: 69bbbf8c47f34ce924a5e531.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### ExpThink

Intelligence Per Token (IPT): 32.84. Result ID: 6a01a40447f34ce924e3e772.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

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

### LC-R1

Intelligence Per Token (IPT): 28.4. Result ID: 69bbbf8c47f34ce924a5e533.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b4a1e209131cea87904.md>)

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

### Laser

Intelligence Per Token (IPT): 27.8. Result ID: 69bbbf8c47f34ce924a5e534.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78b91e209131cea7a10e.md>)

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

### LC-R1

Intelligence Per Token (IPT): 23.47. Result ID: 6a01a40447f34ce924e3e76d.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b4a1e209131cea87904.md>)

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

### Laser

Intelligence Per Token (IPT): 23.3. Result ID: 6a01a40447f34ce924e3e76e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78b91e209131cea7a10e.md>)

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

### ExpThink

Intelligence Per Token (IPT): 23.29. Result ID: 6a01a40447f34ce924e3e76b.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### JET

Intelligence Per Token (IPT): 21.1. Result ID: 69bbbf8c47f34ce924a5e537.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b34f.md>)

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

### AdaptThink

Intelligence Per Token (IPT): 19.52. Result ID: 6a01a40447f34ce924e3e770.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a61e209131cea799d4.md>)

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

### AdaptThink

Intelligence Per Token (IPT): 19.3. Result ID: 69bbbf8c47f34ce924a5e536.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a61e209131cea799d4.md>)

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

### AutoThink

Intelligence Per Token (IPT): 18.63. Result ID: 6a01a40447f34ce924e3e76f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c011e209131cea8bd2d.md>)

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

### AdaptThink

Intelligence Per Token (IPT): 18.5. Result ID: 69bbbf8c47f34ce924a5e52f.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a61e209131cea799d4.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### AdaptThink

Intelligence Per Token (IPT): 18.04. Result ID: 6a01a40447f34ce924e3e769.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a61e209131cea799d4.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### AutoThink

Intelligence Per Token (IPT): 17.9. Result ID: 69bbbf8c47f34ce924a5e535.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c011e209131cea8bd2d.md>)

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

### JET

Intelligence Per Token (IPT): 16.7. Result ID: 69bbbf8c47f34ce924a5e530.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b34f.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### Laser

Intelligence Per Token (IPT): 15.8. Result ID: 69bbbf8c47f34ce924a5e52d.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78b91e209131cea7a10e.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### AutoThink

Intelligence Per Token (IPT): 14.9. Result ID: 69bbbf8c47f34ce924a5e52e.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c011e209131cea8bd2d.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### LC-R1

Intelligence Per Token (IPT): 14.49. Result ID: 6a01a40447f34ce924e3e766.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b4a1e209131cea87904.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### JET

Intelligence Per Token (IPT): 14.35. Result ID: 6a01a40447f34ce924e3e771.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b34f.md>)

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

### LC-R1

Intelligence Per Token (IPT): 13.6. Result ID: 69bbbf8c47f34ce924a5e52c.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b4a1e209131cea87904.md>)

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

### Laser

Intelligence Per Token (IPT): 13.35. Result ID: 6a01a40447f34ce924e3e767.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78b91e209131cea7a10e.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### AutoThink

Intelligence Per Token (IPT): 13.25. Result ID: 6a01a40447f34ce924e3e768.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c011e209131cea8bd2d.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### Vanilla

Intelligence Per Token (IPT): 11.2. Result ID: 69bbbf8c47f34ce924a5e532.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

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

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

### Vanilla

Intelligence Per Token (IPT): 11.15. Result ID: 6a01a40447f34ce924e3e76c.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

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

### JET

Intelligence Per Token (IPT): 10.86. Result ID: 6a01a40447f34ce924e3e76a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b34f.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### Vanilla

Intelligence Per Token (IPT): 8.64. Result ID: 6a01a40447f34ce924e3e773.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

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

### Vanilla

Intelligence Per Token (IPT): 7.23. Result ID: 6a01a40447f34ce924e3e765.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162afdbc836b87d823244.md>)

Recorded evaluation configuration: {"Backbone":"DeepSeek-R1-Distill-Qwen-1.5B"}. This record does not assert independent reproduction.

### Vanilla

Intelligence Per Token (IPT): 7.2. Result ID: 69bbbf8c47f34ce924a5e52b.

[Reporting paper](<https://www.sota2.com/research/paper/69bb83200636e5bf4afc1b3e.md>)

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

Recorded evaluation configuration: {"Base Model":"DeepSeek-R1-Distill-Qwen","Model Scale":"1.5B"}. This record does not assert independent reproduction.

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