# Generative Multiple-choice Question Answering on TruthfulQA

Source: [Generative Multiple-choice Question Answering on TruthfulQA](<https://www.sota2.com/research/sota/generative-multiple-choice-question-answering-on-truthfulqa>)

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

Data updated: 2026-02-26T06:05:50.119Z

Evaluation of model performance on the TruthfulQA dataset using generative multiple-choice question answering metrics, including TA Rate, UR Rate, and DA Rate.

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 | Generative Multiple-choice Question Answering |
| Dataset (including any stated split) | TruthfulQA |
| Metric | TA Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [TA Rate](<https://www.sota2.com/research/sota/generative-multiple-choice-question-answering-on-truthfulqa.md?metric=TA%20Rate>): Higher is better; unit: PERCENT.

- [UR Rate](<https://www.sota2.com/research/sota/generative-multiple-choice-question-answering-on-truthfulqa.md?metric=UR%20Rate>): Higher is better; unit: PERCENT.

- [DA Rate](<https://www.sota2.com/research/sota/generative-multiple-choice-question-answering-on-truthfulqa.md?metric=DA%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | TA Rate | Result date | Report paper |
| --- | --- | --- | --- |
| Llama 2-Chat | 76.3 | Not recorded | 68fa74ce1e209131cea63ff7 |
| Mistral-Instruct-v0.2 | 75.4 | Not recorded | 68fa74ce1e209131cea63ff7 |
| Mistral-Instruct-v0.2 + TACS-S | 46.3 | Not recorded | 68fa74ce1e209131cea63ff7 |
| Mistral-Instruct-v0.2 + TACS-T | 44.9 | Not recorded | 68fa74ce1e209131cea63ff7 |
| Llama 2-Chat + TACS-S | 43.7 | Not recorded | 68fa74ce1e209131cea63ff7 |
| Llama 2-Chat + TACS-T | 43.4 | Not recorded | 68fa74ce1e209131cea63ff7 |

### Llama 2-Chat

TA Rate: 76.3. Result ID: 69937c2e3a6c3148491646d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa701f1e209131cea492fe.md>)

[Implementation](<https://github.com/facebookresearch/llama>)

Recorded evaluation configuration: {"Base Model":"Llama 2-Chat","Selection Strategy":"None"}. This record does not assert independent reproduction.

### Mistral-Instruct-v0.2

TA Rate: 75.4. Result ID: 69937c2e3a6c3148491646d4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d61e209131cea4d909.md>)

[Implementation](<https://github.com/mistralai/mistral-src>)

Recorded evaluation configuration: {"Base Model":"Mistral-Instruct-v0.2","Selection Strategy":"None"}. This record does not assert independent reproduction.

### Mistral-Instruct-v0.2 + TACS-S

TA Rate: 46.3. Result ID: 69937c2e3a6c3148491646d6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Implementation](<https://github.com/ictnlp/TACS>)

Recorded evaluation configuration: {"Base Model":"Mistral-Instruct-v0.2","Selection Strategy":"TACS-S"}. This record does not assert independent reproduction.

### Mistral-Instruct-v0.2 + TACS-T

TA Rate: 44.9. Result ID: 69937c2e3a6c3148491646d5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Implementation](<https://github.com/ictnlp/TACS>)

Recorded evaluation configuration: {"Base Model":"Mistral-Instruct-v0.2","Selection Strategy":"TACS-T"}. This record does not assert independent reproduction.

### Llama 2-Chat + TACS-S

TA Rate: 43.7. Result ID: 69937c2e3a6c3148491646d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Implementation](<https://github.com/ictnlp/TACS>)

Recorded evaluation configuration: {"Base Model":"Llama 2-Chat","Selection Strategy":"TACS-S"}. This record does not assert independent reproduction.

### Llama 2-Chat + TACS-T

TA Rate: 43.4. Result ID: 69937c2e3a6c3148491646d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ce1e209131cea63ff7.md>)

[Implementation](<https://github.com/ictnlp/TACS>)

Recorded evaluation configuration: {"Base Model":"Llama 2-Chat","Selection Strategy":"TACS-T"}. This record does not assert independent reproduction.

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