# Persona-based Role-Playing Faithfulness Evaluation on Complicated Famous Figures (test)

Source: [Persona-based Role-Playing Faithfulness Evaluation on Complicated Famous Figures (test)](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test>)

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

Data updated: 2026-02-26T06:04:32.191Z

Evaluation of Persona-based Role-Playing Faithfulness on the test split of the Complicated Famous Figures dataset, using specific character adherence metrics (ΔAPC) for figures like Spartacus, Hermione, and Caesar.

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 | Persona-based Role-Playing Faithfulness Evaluation |
| Dataset (including any stated split) | Complicated Famous Figures (test) |
| Metric | Spartacus ΔAPC |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 8 of 8 |
| 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

- [Spartacus ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Spartacus%20%CE%94APC>): Lower is better; unit: SCALAR.

- [Hermione ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Hermione%20%CE%94APC>): Lower is better; unit: SCALAR.

- [Voldemort ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Voldemort%20%CE%94APC>): Lower is better; unit: SCALAR.

- [Cleopatra ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Cleopatra%20%CE%94APC>): Lower is better; unit: SCALAR.

- [Caesar ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Caesar%20%CE%94APC>): Lower is better; unit: SCALAR.

- [MLK ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=MLK%20%CE%94APC>): Lower is better; unit: SCALAR.

- [Avg ΔAPC](<https://www.sota2.com/research/sota/persona-based-role-playing-faithfulness-evaluation-on-complicated-famous-figures-test.md?metric=Avg%20%CE%94APC>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Spartacus ΔAPC | Result date | Report paper |
| --- | --- | --- | --- |
| EU | 0.9 | Not recorded | 68fa753e1e209131cea66b68 |
| EU + APC-based DPO | 2.2 | Not recorded | 68fa753e1e209131cea66b68 |
| Gemma-7B | 2.3 | Not recorded | 68fa753e1e209131cea66b68 |
| GPT-4 | 2.5 | Not recorded | 68fa753e1e209131cea66b68 |
| ChatGPT | 2.6 | Not recorded | 68fa753e1e209131cea66b68 |
| Gemma-7B + APC-based DPO | 2.9 | Not recorded | 68fa753e1e209131cea66b68 |
| RAG + APC-based DPO | 3.4 | Not recorded | 68fa753e1e209131cea66b68 |
| RAG | 3.6 | Not recorded | 68fa753e1e209131cea66b68 |

### EU

Spartacus ΔAPC: 0.9. Result ID: 699397123a6c314849171683.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Experience Uploading","APC-based DPO":false}. This record does not assert independent reproduction.

### EU + APC-based DPO

Spartacus ΔAPC: 2.2. Result ID: 699397123a6c314849171686.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Experience Uploading","APC-based DPO":true}. This record does not assert independent reproduction.

### Gemma-7B

Spartacus ΔAPC: 2.3. Result ID: 699397123a6c314849171682.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Vanilla","APC-based DPO":false}. This record does not assert independent reproduction.

### GPT-4

Spartacus ΔAPC: 2.5. Result ID: 699397123a6c314849171681.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"Model Family":"GPT","PRP Strategy":"Vanilla","APC-based DPO":false}. This record does not assert independent reproduction.

### ChatGPT

Spartacus ΔAPC: 2.6. Result ID: 699397123a6c314849171680.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

Recorded evaluation configuration: {"Model Family":"GPT","PRP Strategy":"Vanilla","APC-based DPO":false}. This record does not assert independent reproduction.

### Gemma-7B + APC-based DPO

Spartacus ΔAPC: 2.9. Result ID: 699397123a6c314849171685.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Vanilla","APC-based DPO":true}. This record does not assert independent reproduction.

### RAG + APC-based DPO

Spartacus ΔAPC: 3.4. Result ID: 699397123a6c314849171687.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Retrieval-Augmented Generation","APC-based DPO":true}. This record does not assert independent reproduction.

### RAG

Spartacus ΔAPC: 3.6. Result ID: 699397123a6c314849171684.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753e1e209131cea66b68.md>)

[Implementation](<https://github.com/KomeijiForce/Active_Passive_Constraint_Koishiday_2024>)

Recorded evaluation configuration: {"Model Family":"Gemma-7B","PRP Strategy":"Retrieval-Augmented Generation","APC-based DPO":false}. This record does not assert independent reproduction.

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