# Graph Question Answering on SceneGraphs (test) Source: [Graph Question Answering on SceneGraphs (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:33:19.461Z Evaluation of model performance on the test split of the SceneGraphs dataset for Graph Question Answering tasks, measured using 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 | Graph Question Answering | | Dataset (including any stated split) | SceneGraphs (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 13 of 13 | | 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](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | TopoRAG w/ LoRA | 87.68 | Not recorded | 699d307d6017c4460ffab3d7 | | G-Retriever w/ LoRA | 86.83 | Not recorded | 699d307d6017c4460ffab3d7 | | TopoRAG | 83.62 | Not recorded | 699d307d6017c4460ffab3d7 | | GNN-RAG | 81.49 | Not recorded | 699d307d6017c4460ffab3d7 | | G-Retriever | 81.31 | Not recorded | 699d307d6017c4460ffab3d7 | | SubgraphRAG | 80.74 | Not recorded | 699d307d6017c4460ffab3d7 | | LoRA | 78.62 | Not recorded | 699d307d6017c4460ffab3d7 | | Prompt tuning | 63.41 | Not recorded | 699d307d6017c4460ffab3d7 | | CoT-BAG | 56.8 | Not recorded | 699d307d6017c4460ffab3d7 | | Zero-CoT | 52.6 | Not recorded | 699d307d6017c4460ffab3d7 | | GraphToken | 49.03 | Not recorded | 699d307d6017c4460ffab3d7 | | KAPING | 43.75 | Not recorded | 699d307d6017c4460ffab3d7 | | Zero-shot | 39.74 | Not recorded | 699d307d6017c4460ffab3d7 | ### TopoRAG w/ LoRA Accuracy: 87.68. Result ID: 699e3c0a5698665ec4f8cffd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Tuned LLM"}. This record does not assert independent reproduction. ### G-Retriever w/ LoRA Accuracy: 86.83. Result ID: 699e3c0a5698665ec4f8cffb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Setting":"Tuned LLM"}. This record does not assert independent reproduction. ### TopoRAG Accuracy: 83.62. Result ID: 699e3c0a5698665ec4f8cff9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Frozen LLM w/ PT"}. This record does not assert independent reproduction. ### GNN-RAG Accuracy: 81.49. Result ID: 699e3c0a5698665ec4f8cffc. [Reporting paper]() Recorded evaluation configuration: {"Setting":"Tuned LLM"}. This record does not assert independent reproduction. ### G-Retriever Accuracy: 81.31. Result ID: 699e3c0a5698665ec4f8cff7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Setting":"Frozen LLM w/ PT"}. This record does not assert independent reproduction. ### SubgraphRAG Accuracy: 80.74. Result ID: 699e3c0a5698665ec4f8cff8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Frozen LLM w/ PT"}. This record does not assert independent reproduction. ### LoRA Accuracy: 78.62. Result ID: 699e3c0a5698665ec4f8cffa. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Setting":"Tuned LLM"}. This record does not assert independent reproduction. ### Prompt tuning Accuracy: 63.41. Result ID: 699e3c0a5698665ec4f8cff5. [Reporting paper]() Recorded evaluation configuration: {"Setting":"Frozen LLM w/ PT"}. This record does not assert independent reproduction. ### CoT-BAG Accuracy: 56.8. Result ID: 699e3c0a5698665ec4f8cff3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Inference-only"}. This record does not assert independent reproduction. ### Zero-CoT Accuracy: 52.6. Result ID: 699e3c0a5698665ec4f8cff2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Setting":"Inference-only"}. This record does not assert independent reproduction. ### GraphToken Accuracy: 49.03. Result ID: 699e3c0a5698665ec4f8cff6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Frozen LLM w/ PT"}. This record does not assert independent reproduction. ### KAPING Accuracy: 43.75. Result ID: 699e3c0a5698665ec4f8cff4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setting":"Inference-only"}. This record does not assert independent reproduction. ### Zero-shot Accuracy: 39.74. Result ID: 699e3c0a5698665ec4f8cff1. [Reporting paper]() Recorded evaluation configuration: {"Setting":"Inference-only"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.