# Multi-hop Question Answering on HotpotQA (Size-Cut Evaluation) Source: [Multi-hop Question Answering on HotpotQA (Size-Cut Evaluation)]() Collection index: [llms.txt]() Data updated: 2026-06-04T08:18:35.224Z Evaluation of model performance on the HotpotQA dataset for multi-hop question answering, reporting scores across specific data subset sizes (7k, 14k, ..., 896k) and an overall average. 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-hop Question Answering | | Dataset (including any stated split) | HotpotQA | | Metric | Score (7k) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 8 of 8 | | 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 - [Score (7k)](): Higher is better; unit: PERCENT. - [Score (14k)](): Higher is better; unit: PERCENT. - [Score (28k)](): Higher is better; unit: PERCENT. - [Score (56k)](): Higher is better; unit: PERCENT. - [Score (112k)](): Higher is better; unit: PERCENT. - [Score (224k)](): Higher is better; unit: PERCENT. - [Score (448k)](): Higher is better; unit: PERCENT. - [Score (896k)](): Higher is better; unit: PERCENT. - [Average Score (Overall)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Score (7k) | Result date | Report paper | | --- | --- | --- | --- | | Qwen3-4B-Instruct + MemTrain + MemAgent | 79.69 | Not recorded | 6a21065ec10484bac23efebf | | Qwen2.5-7B-Instruct + MemTrain + MemAgent | 76.56 | Not recorded | 6a21065ec10484bac23efebf | | Qwen3-4B-Instruct + MemAgent | 70.31 | Not recorded | 6a21065ec10484bac23efebf | | Qwen2.5-7B-Instruct + MemAgent | 64.06 | Not recorded | 6a21065ec10484bac23efebf | | Qwen3-4B-Instruct + MemTrain | 63.28 | Not recorded | 6a21065ec10484bac23efebf | | Qwen2.5-7B-Instruct + MemTrain | 59.38 | Not recorded | 6a21065ec10484bac23efebf | | Qwen3-4B-Instruct | 57.81 | Not recorded | 6a21065ec10484bac23efebf | | Qwen2.5-7B-Instruct | 53.12 | Not recorded | 6a21065ec10484bac23efebf | ### Qwen3-4B-Instruct + MemTrain + MemAgent Score (7k): 79.69. Result ID: 6a213331fbc373592c1ca457. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Instruct","Training Stage":"MemTrain + MemAgent"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + MemTrain + MemAgent Score (7k): 76.56. Result ID: 6a213331fbc373592c1ca45b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Training Stage":"MemTrain + MemAgent"}. This record does not assert independent reproduction. ### Qwen3-4B-Instruct + MemAgent Score (7k): 70.31. Result ID: 6a213331fbc373592c1ca456. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Instruct","Training Stage":"MemAgent"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + MemAgent Score (7k): 64.06. Result ID: 6a213331fbc373592c1ca45a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Training Stage":"MemAgent"}. This record does not assert independent reproduction. ### Qwen3-4B-Instruct + MemTrain Score (7k): 63.28. Result ID: 6a213331fbc373592c1ca455. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Instruct","Training Stage":"MemTrain"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + MemTrain Score (7k): 59.38. Result ID: 6a213331fbc373592c1ca459. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Training Stage":"MemTrain"}. This record does not assert independent reproduction. ### Qwen3-4B-Instruct Score (7k): 57.81. Result ID: 6a213331fbc373592c1ca454. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct Score (7k): 53.12. Result ID: 6a213331fbc373592c1ca458. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.