# Process Verification on ProcessBench Without Standard Answers

Source: [Process Verification on ProcessBench Without Standard Answers](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers>)

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

Data updated: 2026-02-26T05:58:11.586Z

Evaluation of model performance on the ProcessBench variant lacking standard answers, focusing on process verification using precise, approximate, and rough accuracy/F1 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 | Process Verification |
| Dataset (including any stated split) | ProcessBench Without Standard Answers |
| Metric | Precise 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

- [Precise Accuracy](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Precise%20Accuracy>): Higher is better; unit: PERCENT.

- [Absolute Accuracy](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Absolute%20Accuracy>): Higher is better; unit: PERCENT.

- [Precise F1 Score](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Precise%20F1%20Score>): Higher is better; unit: PERCENT.

- [Absolute F1 Score](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Absolute%20F1%20Score>): Higher is better; unit: PERCENT.

- [Approximate Accuracy](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Approximate%20Accuracy>): Higher is better; unit: PERCENT.

- [Approximate F1 Score](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Approximate%20F1%20Score>): Higher is better; unit: PERCENT.

- [Rough Accuracy](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Rough%20Accuracy>): Higher is better; unit: PERCENT.

- [Rough F1 Score](<https://www.sota2.com/research/sota/process-verification-on-processbench-without-standard-answers.md?metric=Rough%20F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Precise Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| OPV-32B | 71.9 | Not recorded | 693ba1796cfd488761a55186 |
| OPV-Stage2 | 64.4 | Not recorded | 693ba1796cfd488761a55186 |
| R1-Distill-Qwen-32B | 61.7 | Not recorded | 693ba1796cfd488761a55186 |
| Qwen3-32B | 61.2 | Not recorded | 693ba1796cfd488761a55186 |
| Qwen3-Max-Preview | 61 | Not recorded | 693ba1796cfd488761a55186 |
| OPV-Stage1 | 60.9 | Not recorded | 693ba1796cfd488761a55186 |
| DeepSeek-V3-0324 | 60.8 | Not recorded | 693ba1796cfd488761a55186 |
| QwQ-32B | 58.2 | Not recorded | 693ba1796cfd488761a55186 |
| gpt-oss-120b (high) | 57.9 | Not recorded | 693ba1796cfd488761a55186 |
| DeepSeek-R1-0528 | 56.9 | Not recorded | 693ba1796cfd488761a55186 |
| Qwen2.5-Math-PRM-72B | 55.1 | Not recorded | 693ba1796cfd488761a55186 |
| gpt-oss-120b (high) | 0.847 | Not recorded | 693ba1796cfd488761a55186 |
| Qwen3-Max-Preview | 0.844 | Not recorded | 693ba1796cfd488761a55186 |
| DeepSeek-R1-0528 | 0.832 | Not recorded | 693ba1796cfd488761a55186 |
| OPV-32B | 0.809 | Not recorded | 693ba1796cfd488761a55186 |
| Qwen2.5-Math-PRM-72B | 0.767 | Not recorded | 693ba1796cfd488761a55186 |
| R1-Distill-Qwen-32B | 0.758 | Not recorded | 693ba1796cfd488761a55186 |
| DeepSeek-V3-0324 | 0.721 | Not recorded | 693ba1796cfd488761a55186 |

### OPV-32B

Precise Accuracy: 71.9. Result ID: 6996c36f5698665ec40ec530.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### OPV-Stage2

Precise Accuracy: 64.4. Result ID: 6996c36f5698665ec40ec52f.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### R1-Distill-Qwen-32B

Precise Accuracy: 61.7. Result ID: 6996c36f5698665ec40ec52d.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### Qwen3-32B

Precise Accuracy: 61.2. Result ID: 6996c36f5698665ec40ec526.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### Qwen3-Max-Preview

Precise Accuracy: 61. Result ID: 6996c36f5698665ec40ec52a.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### OPV-Stage1

Precise Accuracy: 60.9. Result ID: 6996c36f5698665ec40ec52e.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### DeepSeek-V3-0324

Precise Accuracy: 60.8. Result ID: 6996c36f5698665ec40ec528.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77001e209131cea7066a.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### QwQ-32B

Precise Accuracy: 58.2. Result ID: 6996c36f5698665ec40ec527.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### gpt-oss-120b (high)

Precise Accuracy: 57.9. Result ID: 6996c36f5698665ec40ec52b.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea84a0c.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### DeepSeek-R1-0528

Precise Accuracy: 56.9. Result ID: 6996c36f5698665ec40ec529.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### Qwen2.5-Math-PRM-72B

Precise Accuracy: 55.1. Result ID: 6996c36f5698665ec40ec52c.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### gpt-oss-120b (high)

Precise Accuracy: 0.847. Result ID: 6996c36f5698665ec40ec517.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea84a0c.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### Qwen3-Max-Preview

Precise Accuracy: 0.844. Result ID: 6996c36f5698665ec40ec514.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### DeepSeek-R1-0528

Precise Accuracy: 0.832. Result ID: 6996c36f5698665ec40ec516.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### OPV-32B

Precise Accuracy: 0.809. Result ID: 6996c36f5698665ec40ec51a.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### Qwen2.5-Math-PRM-72B

Precise Accuracy: 0.767. Result ID: 6996c36f5698665ec40ec518.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### R1-Distill-Qwen-32B

Precise Accuracy: 0.758. Result ID: 6996c36f5698665ec40ec519.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

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

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

### DeepSeek-V3-0324

Precise Accuracy: 0.721. Result ID: 6996c36f5698665ec40ec515.

[Reporting paper](<https://www.sota2.com/research/paper/693ba1796cfd488761a55186.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77001e209131cea7066a.md>)

Recorded evaluation configuration: {"Voting":"maj@8"}. This record does not assert independent reproduction.

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