# Multi-document Retrieval on HotpotQA, 2WikiMultihopQA, and MuSiQue (val) Source: [Multi-document Retrieval on HotpotQA, 2WikiMultihopQA, and MuSiQue (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:35:26.320Z Evaluation of multi-document retrieval performance across the validation splits of HotpotQA, 2WikiMultihopQA, and MuSiQue datasets, measured using Average F-1 score. 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 | Multi-document retrieval | | Dataset (including any stated split) | HotpotQA, 2WikiMultihopQA, and MuSiQue (val) | | Metric | Average F1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 11 of 11 | | 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 - [Average F1](): Higher is better; unit: PERCENT. ## Recorded results | Method | Average F1 | Result date | Report paper | | --- | --- | --- | --- | | GritLM | 64.61 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | GTE-Qwen2 | 62.25 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | AttentionRetriever | 62.23 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | Qwen3 | 61.58 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | AttentionRetriever | 61.51 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | GTR | 56.98 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | ANCE | 54.78 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | SPScanner | 54.34 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | BM25 | 48.54 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | DPR | 46.96 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | | CDE | 29.63 | Not recorded | 698eb05dc97a0b3e3b5e69a4 | ### GritLM Average F1: 64.61. Result ID: 69989ceb5698665ec4ff45ba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### GTE-Qwen2 Average F1: 62.25. Result ID: 69989ceb5698665ec4ff45b8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### AttentionRetriever Average F1: 62.23. Result ID: 69989ceb5698665ec4ff45bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"AttentionRetriever (Ours)","Backbone":"LLaMA-3.2","Model size":"3B","top\_k":3}. This record does not assert independent reproduction. ### Qwen3 Average F1: 61.58. Result ID: 69989ceb5698665ec4ff45b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### AttentionRetriever Average F1: 61.51. Result ID: 69989ceb5698665ec4ff45bd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"AttentionRetriever (Ours)","Backbone":"Qwen-2.5","Model size":"3B","top\_k":3}. This record does not assert independent reproduction. ### GTR Average F1: 56.98. Result ID: 69989ceb5698665ec4ff45b7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### ANCE Average F1: 54.78. Result ID: 69989ceb5698665ec4ff45b5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### SPScanner Average F1: 54.34. Result ID: 69989ceb5698665ec4ff45bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Autoregressive Models","top\_k":3}. This record does not assert independent reproduction. ### BM25 Average F1: 48.54. Result ID: 69989ceb5698665ec4ff45b3. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Sparse Models","top\_k":3}. This record does not assert independent reproduction. ### DPR Average F1: 46.96. Result ID: 69989ceb5698665ec4ff45b4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. ### CDE Average F1: 29.63. Result ID: 69989ceb5698665ec4ff45b6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Dense Models","top\_k":3}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.