# Multi-turn Refinement on DeepResearch Bench 100 follow-up instances

Source: [Multi-turn Refinement on DeepResearch Bench 100 follow-up instances](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances>)

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

Data updated: 2026-06-19T08:23:38.256Z

Evaluation of model performance on the Multi-turn Refinement task using the DeepResearch Bench 100 follow-up instances dataset, measured by component scores (Instance, Evidence, Logic, Completeness, Explanation) and a Total score.

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 | Multi-turn Refinement |
| Dataset (including any stated split) | DeepResearch Bench 100 follow-up instances |
| Metric | Instance Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [Instance Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Instance%20Score>): Higher is better; unit: SCALAR.

- [Evidence Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Evidence%20Score>): Higher is better; unit: SCALAR.

- [Logic Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Logic%20Score>): Higher is better; unit: SCALAR.

- [Completeness Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Completeness%20Score>): Higher is better; unit: SCALAR.

- [Explanation Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Explanation%20Score>): Higher is better; unit: SCALAR.

- [Total Score](<https://www.sota2.com/research/sota/multi-turn-refinement-on-deepresearch-bench-100-follow-up-instances.md?metric=Total%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Instance Score | Result date | Report paper |
| --- | --- | --- | --- |
| SCAFFOLDAGENT | 15.35 | Not recorded | 6a34cd4b206542251d7c6185 |
| ReAct-FW | 13.8 | Not recorded | 6a34cd4b206542251d7c6185 |
| ReAct-GR | 12.65 | Not recorded | 6a34cd4b206542251d7c6185 |

### SCAFFOLDAGENT

Instance Score: 15.35. Result ID: 6a34fc0efbc373592c324a4b.

[Reporting paper](<https://www.sota2.com/research/paper/6a34cd4b206542251d7c6185.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a34cd4b206542251d7c6185.md>)

Recorded evaluation configuration: {"Setting":"SCAFFOLDAGENT"}. This record does not assert independent reproduction.

### ReAct-FW

Instance Score: 13.8. Result ID: 6a34fc0efbc373592c324a49.

[Reporting paper](<https://www.sota2.com/research/paper/6a34cd4b206542251d7c6185.md>)

Recorded evaluation configuration: {"Setting":"ReAct-FW"}. This record does not assert independent reproduction.

### ReAct-GR

Instance Score: 12.65. Result ID: 6a34fc0efbc373592c324a4a.

[Reporting paper](<https://www.sota2.com/research/paper/6a34cd4b206542251d7c6185.md>)

Recorded evaluation configuration: {"Setting":"ReAct-GR"}. This record does not assert independent reproduction.

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