# Multimodal Question Answering on WebQA Text-grounded n=2,455 (test) Source: [Multimodal Question Answering on WebQA Text-grounded n=2,455 (test)]() Collection index: [llms.txt]() Data updated: 2026-06-11T08:05:28.332Z Evaluation of model performance on the specific test split (n=2,455, text-grounded subset) of the WebQA dataset for Multimodal Question Answering, reporting F1, Accuracy, and R@k. 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 | Multimodal Question Answering | | Dataset (including any stated split) | WebQA Text-grounded n=2,455 (test) | | Metric | F1 Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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](): Higher is better; unit: PERCENT. - [Accuracy](): Higher is better; unit: PERCENT. - [Recall@k](): Higher is better; unit: PERCENT. ## Recorded results | Method | F1 Score | Result date | Report paper | | --- | --- | --- | --- | | Nemo-Emb-1B | 48.6 | Not recorded | 6a28f36e90966ade919c18ff | | Qwen3-VL-Emb-8B | 47.8 | Not recorded | 6a28f36e90966ade919c18ff | | Nemo-Emb-1B | 46.8 | Not recorded | 6a28f36e90966ade919c18ff | | BM25 Retrieval | 46.7 | Not recorded | 6a28f36e90966ade919c18ff | | Qwen3-VL-Emb-8B | 46.1 | Not recorded | 6a28f36e90966ade919c18ff | | BM25 Retrieval | 43.9 | Not recorded | 6a28f36e90966ade919c18ff | | Nemo-Emb-1B | 41.1 | Not recorded | 6a28f36e90966ade919c18ff | | Qwen3-VL-Emb-8B | 40.8 | Not recorded | 6a28f36e90966ade919c18ff | | BM25 Retrieval | 39.9 | Not recorded | 6a28f36e90966ade919c18ff | | Dense Retrieval | 38.5 | Not recorded | 6a28f36e90966ade919c18ff | | Dense Retrieval | 37.6 | Not recorded | 6a28f36e90966ade919c18ff | | Dense Retrieval | 34.9 | Not recorded | 6a28f36e90966ade919c18ff | | Latent Memory | 30.7 | Not recorded | 6a28f36e90966ade919c18ff | | Latent Memory | 30 | Not recorded | 6a28f36e90966ade919c18ff | | Latent Memory | 28.8 | Not recorded | 6a28f36e90966ade919c18ff | | Full Context | 6 | Not recorded | 6a28f36e90966ade919c18ff | ### Nemo-Emb-1B F1 Score: 48.6. Result ID: 6a2a6ae0fbc373592c275357. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"5","Generator VLM":"LLaVA-1.5-13B","#Tok":"629"}. This record does not assert independent reproduction. ### Qwen3-VL-Emb-8B F1 Score: 47.8. Result ID: 6a2a6ae0fbc373592c27535a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"5","Generator VLM":"LLaVA-1.5-13B","#Tok":"612"}. This record does not assert independent reproduction. ### Nemo-Emb-1B F1 Score: 46.8. Result ID: 6a2a6ae0fbc373592c275356. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"2","Generator VLM":"LLaVA-1.5-13B","#Tok":"233"}. This record does not assert independent reproduction. ### BM25 Retrieval F1 Score: 46.7. Result ID: 6a2a6ae0fbc373592c275351. [Reporting paper]() Recorded evaluation configuration: {"k":"5","Generator VLM":"LLaVA-1.5-13B","#Tok":"552"}. This record does not assert independent reproduction. ### Qwen3-VL-Emb-8B F1 Score: 46.1. Result ID: 6a2a6ae0fbc373592c275359. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"2","Generator VLM":"LLaVA-1.5-13B","#Tok":"235"}. This record does not assert independent reproduction. ### BM25 Retrieval F1 Score: 43.9. Result ID: 6a2a6ae0fbc373592c275350. [Reporting paper]() Recorded evaluation configuration: {"k":"2","Generator VLM":"LLaVA-1.5-13B","#Tok":"234"}. This record does not assert independent reproduction. ### Nemo-Emb-1B F1 Score: 41.1. Result ID: 6a2a6ae0fbc373592c275355. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"1","Generator VLM":"LLaVA-1.5-13B","#Tok":"130"}. This record does not assert independent reproduction. ### Qwen3-VL-Emb-8B F1 Score: 40.8. Result ID: 6a2a6ae0fbc373592c275358. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"1","Generator VLM":"LLaVA-1.5-13B","#Tok":"131"}. This record does not assert independent reproduction. ### BM25 Retrieval F1 Score: 39.9. Result ID: 6a2a6ae0fbc373592c27534f. [Reporting paper]() Recorded evaluation configuration: {"k":"1","Generator VLM":"LLaVA-1.5-13B","#Tok":"133"}. This record does not assert independent reproduction. ### Dense Retrieval F1 Score: 38.5. Result ID: 6a2a6ae0fbc373592c275353. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"2","Generator VLM":"LLaVA-1.5-13B","#Tok":"538"}. This record does not assert independent reproduction. ### Dense Retrieval F1 Score: 37.6. Result ID: 6a2a6ae0fbc373592c275354. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"5","Generator VLM":"LLaVA-1.5-13B","#Tok":"1396"}. This record does not assert independent reproduction. ### Dense Retrieval F1 Score: 34.9. Result ID: 6a2a6ae0fbc373592c275352. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"1","Generator VLM":"LLaVA-1.5-13B","#Tok":"264"}. This record does not assert independent reproduction. ### Latent Memory F1 Score: 30.7. Result ID: 6a2a6ae0fbc373592c27535d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"5","Generator VLM":"LLaVA-1.5-13B","#Tok":"84"}. This record does not assert independent reproduction. ### Latent Memory F1 Score: 30. Result ID: 6a2a6ae0fbc373592c27535c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"2","Generator VLM":"LLaVA-1.5-13B","#Tok":"54"}. This record does not assert independent reproduction. ### Latent Memory F1 Score: 28.8. Result ID: 6a2a6ae0fbc373592c27535b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"k":"1","Generator VLM":"LLaVA-1.5-13B","#Tok":"44"}. This record does not assert independent reproduction. ### Full Context F1 Score: 6. Result ID: 6a2a6ae0fbc373592c27534e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generator VLM":"LLaVA-1.5-13B","#Tok":"8371"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.