# Personalized Question Answering on LaMP-QA

Source: [Personalized Question Answering on LaMP-QA](<https://www.sota2.com/research/sota/personalized-question-answering-on-lamp-qa>)

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

Data updated: 2026-02-26T05:34:09.075Z

Evaluation of model performance on the LaMP-QA dataset for Personalized Question Answering, reporting category-specific accuracy scores.

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 | Personalized Question Answering |
| Dataset (including any stated split) | LaMP-QA |
| Metric | Accuracy (Arts & Entertainment) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [Accuracy (Arts & Entertainment)](<https://www.sota2.com/research/sota/personalized-question-answering-on-lamp-qa.md?metric=Accuracy%20(Arts%20%26%20Entertainment)>): Higher is better; unit: PERCENT.

- [Accuracy (Lifestyle & Personal Development)](<https://www.sota2.com/research/sota/personalized-question-answering-on-lamp-qa.md?metric=Accuracy%20(Lifestyle%20%26%20Personal%20Development)>): Higher is better; unit: PERCENT.

- [Accuracy (Society & Culture)](<https://www.sota2.com/research/sota/personalized-question-answering-on-lamp-qa.md?metric=Accuracy%20(Society%20%26%20Culture)>): Higher is better; unit: PERCENT.

- [Accuracy (Macro Average)](<https://www.sota2.com/research/sota/personalized-question-answering-on-lamp-qa.md?metric=Accuracy%20(Macro%20Average)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (Arts & Entertainment) | Result date | Report paper |
| --- | --- | --- | --- |
| Upper Bound: RAG-Personalization Server-Side | 53.27 | Not recorded | 69784675a2b28e761d87c422 |
| P3 | 48.11 | Not recorded | 69784675a2b28e761d87c422 |
| RAG-Personalization Client-Side | 42.58 | Not recorded | 69784675a2b28e761d87c422 |
| Non-Personalized Server-Side | 41.32 | Not recorded | 69784675a2b28e761d87c422 |
| Upper Bound: RAG-Personalization Server-Side | 39.2 | Not recorded | 69784675a2b28e761d87c422 |
| Non-Personalized Speculative Decoding | 39.12 | Not recorded | 69784675a2b28e761d87c422 |
| P3 | 36.8 | Not recorded | 69784675a2b28e761d87c422 |
| Non-Personalized Server-Side | 32.66 | Not recorded | 69784675a2b28e761d87c422 |
| Non-Personalized Speculative Decoding | 32.26 | Not recorded | 69784675a2b28e761d87c422 |
| RAG-Personalization Client-Side | 30.13 | Not recorded | 69784675a2b28e761d87c422 |

### Upper Bound: RAG-Personalization Server-Side

Accuracy (Arts & Entertainment): 53.27. Result ID: 69966f9f5698665ec4b69f5f.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Gemma 3","Parameters":"12B"}. This record does not assert independent reproduction.

### P3

Accuracy (Arts & Entertainment): 48.11. Result ID: 69966f9f5698665ec4b69f5e.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Gemma 3","Server-Side Parameters":"12B","Client-Side Parameters":"4B"}. This record does not assert independent reproduction.

### RAG-Personalization Client-Side

Accuracy (Arts & Entertainment): 42.58. Result ID: 69966f9f5698665ec4b69f5d.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73701e209131cea5c7af.md>)

[Implementation](<http://lamp-benchmark.github.io/>)

Recorded evaluation configuration: {"Model Family":"Gemma 3","Parameters":"4B"}. This record does not assert independent reproduction.

### Non-Personalized Server-Side

Accuracy (Arts & Entertainment): 41.32. Result ID: 69966f9f5698665ec4b69f5b.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Gemma 3","Parameters":"12B"}. This record does not assert independent reproduction.

### Upper Bound: RAG-Personalization Server-Side

Accuracy (Arts & Entertainment): 39.2. Result ID: 69966f9f5698665ec4b69f5a.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Qwen 2.5","Parameters":"14B"}. This record does not assert independent reproduction.

### Non-Personalized Speculative Decoding

Accuracy (Arts & Entertainment): 39.12. Result ID: 69966f9f5698665ec4b69f5c.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6f1e209131cea458d8.md>)

Recorded evaluation configuration: {"Model Family":"Gemma 3","Verifier Parameters":"12B","Draft Parameters":"4B"}. This record does not assert independent reproduction.

### P3

Accuracy (Arts & Entertainment): 36.8. Result ID: 69966f9f5698665ec4b69f59.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Qwen 2.5","Server-Side Parameters":"14B","Client-Side Parameters":"3B"}. This record does not assert independent reproduction.

### Non-Personalized Server-Side

Accuracy (Arts & Entertainment): 32.66. Result ID: 69966f9f5698665ec4b69f56.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

Recorded evaluation configuration: {"Model Family":"Qwen 2.5","Parameters":"14B"}. This record does not assert independent reproduction.

### Non-Personalized Speculative Decoding

Accuracy (Arts & Entertainment): 32.26. Result ID: 69966f9f5698665ec4b69f57.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6f1e209131cea458d8.md>)

Recorded evaluation configuration: {"Model Family":"Qwen 2.5","Verifier Parameters":"14B","Draft Parameters":"3B"}. This record does not assert independent reproduction.

### RAG-Personalization Client-Side

Accuracy (Arts & Entertainment): 30.13. Result ID: 69966f9f5698665ec4b69f58.

[Reporting paper](<https://www.sota2.com/research/paper/69784675a2b28e761d87c422.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73701e209131cea5c7af.md>)

[Implementation](<http://lamp-benchmark.github.io/>)

Recorded evaluation configuration: {"Model Family":"Qwen 2.5","Parameters":"3B"}. This record does not assert independent reproduction.

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