# PII Detection on Crash narratives dataset

Source: [PII Detection on Crash narratives dataset](<https://www.sota2.com/research/sota/pii-detection-on-crash-narratives-dataset>)

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

Data updated: 2026-04-20T06:53:51.323Z

Evaluation of model performance on detecting Personally Identifiable Information (PII) within crash narrative texts, measured using standard classification metrics like Precision, Recall, F1-score, and 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 | PII Detection |
| Dataset (including any stated split) | Crash narratives dataset |
| Metric | Precision |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 29 of 29 |
| 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

- [Precision](<https://www.sota2.com/research/sota/pii-detection-on-crash-narratives-dataset.md?metric=Precision>): Higher is better; unit: PERCENT.

- [Recall](<https://www.sota2.com/research/sota/pii-detection-on-crash-narratives-dataset.md?metric=Recall>): Higher is better; unit: PERCENT.

- [F1-score](<https://www.sota2.com/research/sota/pii-detection-on-crash-narratives-dataset.md?metric=F1-score>): Higher is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/pii-detection-on-crash-narratives-dataset.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Precision | Result date | Report paper |
| --- | --- | --- | --- |
| Presidio | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 100 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 98 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 94 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 93 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 91 | Not recorded | 69e5b3027d0172140ec7df1c |
| Presidio | 88 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 88 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 88 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 82 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 78 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 61 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 56 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 54 | Not recorded | 69e5b3027d0172140ec7df1c |
| Agentic Workflow | 54 | Not recorded | 69e5b3027d0172140ec7df1c |
| Presidio | 51 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 50 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 50 | Not recorded | 69e5b3027d0172140ec7df1c |
| Fine-tuned LLM (Llama3.1-8B) | 48 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 41 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 40 | Not recorded | 69e5b3027d0172140ec7df1c |
| Presidio | 38 | Not recorded | 69e5b3027d0172140ec7df1c |
| Gemma-27B prompting(few shot) | 34 | Not recorded | 69e5b3027d0172140ec7df1c |
| Presidio | 8 | Not recorded | 69e5b3027d0172140ec7df1c |
| Llama3.1-8B prompting | 0 | Not recorded | 69e5b3027d0172140ec7df1c |

### Presidio

Precision: 100. Result ID: 69e5cc3f47f34ce924cb65e8.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Email"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 100. Result ID: 69e5cc3f47f34ce924cb65ee.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Alphanumeric"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 100. Result ID: 69e5cc3f47f34ce924cb65f3.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Email"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 100. Result ID: 69e5cc3f47f34ce924cb65f9.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Email"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 100. Result ID: 69e5cc3f47f34ce924cb65ff.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Email"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 100. Result ID: 69e5cc3f47f34ce924cb6601.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Home address"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 98. Result ID: 69e5cc3f47f34ce924cb65f1.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Name"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 94. Result ID: 69e5cc3f47f34ce924cb65f7.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Name"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 93. Result ID: 69e5cc3f47f34ce924cb65fd.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Name"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 91. Result ID: 69e5cc3f47f34ce924cb65fc.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Overall"}. This record does not assert independent reproduction.

### Presidio

Precision: 88. Result ID: 69e5cc3f47f34ce924cb65e7.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Phone Number"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 88. Result ID: 69e5cc3f47f34ce924cb65f2.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Phone Number"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 88. Result ID: 69e5cc3f47f34ce924cb65fe.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Phone Number"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 82. Result ID: 69e5cc3f47f34ce924cb6602.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Overall"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 78. Result ID: 69e5cc3f47f34ce924cb65f6.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Overall"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 61. Result ID: 69e5cc3f47f34ce924cb65f5.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Home address"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 56. Result ID: 69e5cc3f47f34ce924cb65fb.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Home address"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 54. Result ID: 69e5cc3f47f34ce924cb65f8.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Phone Number"}. This record does not assert independent reproduction.

### Agentic Workflow

Precision: 54. Result ID: 69e5cc3f47f34ce924cb6600.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Alphanumeric"}. This record does not assert independent reproduction.

### Presidio

Precision: 51. Result ID: 69e5cc3f47f34ce924cb65e6.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Name"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 50. Result ID: 69e5cc3f47f34ce924cb65ed.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Email"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 50. Result ID: 69e5cc3f47f34ce924cb65ef.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Home address"}. This record does not assert independent reproduction.

### Fine-tuned LLM (Llama3.1-8B)

Precision: 48. Result ID: 69e5cc3f47f34ce924cb65fa.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Alphanumeric"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 41. Result ID: 69e5cc3f47f34ce924cb65f0.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Overall"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 40. Result ID: 69e5cc3f47f34ce924cb65eb.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Name"}. This record does not assert independent reproduction.

### Presidio

Precision: 38. Result ID: 69e5cc3f47f34ce924cb65ea.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Overall"}. This record does not assert independent reproduction.

### Gemma-27B prompting(few shot)

Precision: 34. Result ID: 69e5cc3f47f34ce924cb65f4.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea5973c.md>)

Recorded evaluation configuration: {"PII Category":"Alphanumeric"}. This record does not assert independent reproduction.

### Presidio

Precision: 8. Result ID: 69e5cc3f47f34ce924cb65e9.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

Recorded evaluation configuration: {"PII Category":"Alphanumeric"}. This record does not assert independent reproduction.

### Llama3.1-8B prompting

Precision: 0. Result ID: 69e5cc3f47f34ce924cb65ec.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b3027d0172140ec7df1c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"PII Category":"Phone Number"}. This record does not assert independent reproduction.

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