# Agentic Person Search on Track 1 (Who)

Source: [Agentic Person Search on Track 1 (Who)](<https://www.sota2.com/research/sota/agentic-person-search-on-track-1-who>)

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

Data updated: 2026-04-15T09:44:06.422Z

Evaluation of model performance on the Agentic Person Search task using the Track 1 (Who) dataset, measured by Success Rate (SR@1, SR@5) and Parse Accuracy.

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 | Agentic Person Search |
| Dataset (including any stated split) | Track 1 (Who) |
| Metric | SR@1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [SR@1](<https://www.sota2.com/research/sota/agentic-person-search-on-track-1-who.md?metric=SR%401>): Higher is better; unit: PERCENT.

- [SR@5](<https://www.sota2.com/research/sota/agentic-person-search-on-track-1-who.md?metric=SR%405>): Higher is better; unit: PERCENT.

- [Parse Accuracy](<https://www.sota2.com/research/sota/agentic-person-search-on-track-1-who.md?metric=Parse%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | SR@1 | Result date | Report paper |
| --- | --- | --- | --- |
| Oracle | 100 | Not recorded | 69df1bbf7d0172140ec7bb1b |
| LLM ToolCall | 81.1 | Not recorded | 69df1bbf7d0172140ec7bb1b |
| LLM Direct | 73.3 | Not recorded | 69df1bbf7d0172140ec7bb1b |
| Rule-based | 32.2 | Not recorded | 69df1bbf7d0172140ec7bb1b |

### Oracle

SR@1: 100. Result ID: 69df5dd347f34ce924c882de.

[Reporting paper](<https://www.sota2.com/research/paper/69df1bbf7d0172140ec7bb1b.md>)

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

### LLM ToolCall

SR@1: 81.1. Result ID: 69df5dd347f34ce924c882df.

[Reporting paper](<https://www.sota2.com/research/paper/69df1bbf7d0172140ec7bb1b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69df1bbf7d0172140ec7bb1b.md>)

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

### LLM Direct

SR@1: 73.3. Result ID: 69df5dd347f34ce924c882e0.

[Reporting paper](<https://www.sota2.com/research/paper/69df1bbf7d0172140ec7bb1b.md>)

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

### Rule-based

SR@1: 32.2. Result ID: 69df5dd347f34ce924c882e1.

[Reporting paper](<https://www.sota2.com/research/paper/69df1bbf7d0172140ec7bb1b.md>)

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

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