# Search-based Question Answering on DeepResearchBench (test)

Source: [Search-based Question Answering on DeepResearchBench (test)](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test>)

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

Data updated: 2026-05-25T08:57:39.606Z

Evaluation of Search-based Question Answering performance on the DeepResearchBench (test) split, reporting component scores (Comp., Ins., IF, Read.) and the overall Average.

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 | Search-based Question Answering |
| Dataset (including any stated split) | DeepResearchBench (test) |
| Metric | Component Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 12 of 12 |
| 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

- [Component Score](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test.md?metric=Component%20Score>): Higher is better; unit: PERCENT.

- [Instruction Score](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test.md?metric=Instruction%20Score>): Higher is better; unit: PERCENT.

- [IF Score](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test.md?metric=IF%20Score>): Higher is better; unit: PERCENT.

- [Reading Score](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test.md?metric=Reading%20Score>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/search-based-question-answering-on-deepresearchbench-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Component Score | Result date | Report paper |
| --- | --- | --- | --- |
| Co-ReAct | 34.63 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| CRITIC | 32.53 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Co-ReAct | 31.48 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Self-Refine | 31.37 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| ReAct | 31.25 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Self-Refine | 30.71 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Best-of-N | 30.34 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| CRITIC | 30.25 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Best-of-N | 30.19 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| ReAct | 30.13 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Step-Back | 28.4 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |
| Step-Back | 27.93 | Not recorded | 6a13d7a06f71c5ae7dc9e9aa |

### Co-ReAct

Component Score: 34.63. Result ID: 6a140e69fbc373592c0b602f.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### CRITIC

Component Score: 32.53. Result ID: 6a140e69fbc373592c0b602e.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea5581a.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Co-ReAct

Component Score: 31.48. Result ID: 6a140e69fbc373592c0b6029.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Self-Refine

Component Score: 31.37. Result ID: 6a140e69fbc373592c0b602b.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f7f1e209131cea45fc4.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### ReAct

Component Score: 31.25. Result ID: 6a140e69fbc373592c0b602a.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6eb41e209131cea41992.md>)

[Implementation](<https://react-lm.github.io/>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Self-Refine

Component Score: 30.71. Result ID: 6a140e69fbc373592c0b6025.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f7f1e209131cea45fc4.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Best-of-N

Component Score: 30.34. Result ID: 6a140e69fbc373592c0b602c.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa745f1e209131cea61428.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### CRITIC

Component Score: 30.25. Result ID: 6a140e69fbc373592c0b6028.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea5581a.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Best-of-N

Component Score: 30.19. Result ID: 6a140e69fbc373592c0b6026.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa745f1e209131cea61428.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### ReAct

Component Score: 30.13. Result ID: 6a140e69fbc373592c0b6024.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6eb41e209131cea41992.md>)

[Implementation](<https://react-lm.github.io/>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Step-Back

Component Score: 28.4. Result ID: 6a140e69fbc373592c0b6027.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72761e209131cea56856.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-8B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

### Step-Back

Component Score: 27.93. Result ID: 6a140e69fbc373592c0b602d.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7a06f71c5ae7dc9e9aa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72761e209131cea56856.md>)

Recorded evaluation configuration: {"Search Agent":"Qwen3-14B","Answer Rewriter":"Qwen3-235B"}. This record does not assert independent reproduction.

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