# Knowledge Graph Question Answering on GrailQA IID

Source: [Knowledge Graph Question Answering on GrailQA IID](<https://www.sota2.com/research/sota/knowledge-graph-question-answering-on-grailqa-iid>)

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

Data updated: 2026-02-26T05:43:39.948Z

Evaluation of Knowledge Graph Question Answering performance on the I.I.D. setting of the GrailQA dataset, reporting F1 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 | Knowledge Graph Question Answering |
| Dataset (including any stated split) | GrailQA IID |
| Metric | F1 Score |
| 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

- [F1 Score](<https://www.sota2.com/research/sota/knowledge-graph-question-answering-on-grailqa-iid.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | F1 Score | Result date | Report paper |
| --- | --- | --- | --- |
| PathHD | 92.4 | Not recorded | 693a4fe96cfd488761a548ab |
| KG-Agent | 92 | Not recorded | 693a4fe96cfd488761a548ab |
| StructGPT | 70.4 | Not recorded | 693a4fe96cfd488761a548ab |
| GPT-4 | 25 | Not recorded | 693a4fe96cfd488761a548ab |
| Davinci-003 | 23.5 | Not recorded | 693a4fe96cfd488761a548ab |
| ChatGPT | 19.6 | Not recorded | 693a4fe96cfd488761a548ab |

### PathHD

F1 Score: 92.4. Result ID: 699974a75698665ec4e3220b.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

Recorded evaluation configuration: {"Type":"LLMs + KG"}. This record does not assert independent reproduction.

### KG-Agent

F1 Score: 92. Result ID: 699974a75698665ec4e3220a.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72391e209131cea55147.md>)

Recorded evaluation configuration: {"Type":"LLMs + KG"}. This record does not assert independent reproduction.

### StructGPT

F1 Score: 70.4. Result ID: 699974a75698665ec4e32209.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e44b.md>)

Recorded evaluation configuration: {"Type":"LLMs + KG"}. This record does not assert independent reproduction.

### GPT-4

F1 Score: 25. Result ID: 699974a75698665ec4e32208.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

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

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

Recorded evaluation configuration: {"Type":"Pure LLMs"}. This record does not assert independent reproduction.

### Davinci-003

F1 Score: 23.5. Result ID: 699974a75698665ec4e32207.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

Recorded evaluation configuration: {"Type":"Pure LLMs"}. This record does not assert independent reproduction.

### ChatGPT

F1 Score: 19.6. Result ID: 699974a75698665ec4e32206.

[Reporting paper](<https://www.sota2.com/research/paper/693a4fe96cfd488761a548ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

Recorded evaluation configuration: {"Type":"Pure LLMs"}. This record does not assert independent reproduction.

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