# Question Answering on ZH-4O (test) Source: [Question Answering on ZH-4O (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:59:05.401Z Evaluation of model performance on the test split of the ZH-4O dataset for Question Answering tasks, measured using QA 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 | Question Answering | | Dataset (including any stated split) | ZH-4O (test) | | Metric | QA Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [QA Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | QA Accuracy | Result date | Report paper | | --- | --- | --- | --- | | TELEMEM | 86.33 | Not recorded | 6965d1576343fa78c83232aa | | Long context LLM | 84.92 | Not recorded | 6965d1576343fa78c83232aa | | Memobase | 76.78 | Not recorded | 6965d1576343fa78c83232aa | | A-Mem | 73.78 | Not recorded | 6965d1576343fa78c83232aa | | MOOM | 72.6 | Not recorded | 6965d1576343fa78c83232aa | | Mem0 | 70.2 | Not recorded | 6965d1576343fa78c83232aa | | RAG | 62.45 | Not recorded | 6965d1576343fa78c83232aa | ### TELEMEM QA Accuracy: 86.33. Result ID: 6996c2b95698665ec40e0143. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### Long context LLM QA Accuracy: 84.92. Result ID: 6996c2b95698665ec40e0144. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### Memobase QA Accuracy: 76.78. Result ID: 6996c2b95698665ec40e0145. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### A-Mem QA Accuracy: 73.78. Result ID: 6996c2b95698665ec40e0146. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### MOOM QA Accuracy: 72.6. Result ID: 6996c2b95698665ec40e0147. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### Mem0 QA Accuracy: 70.2. Result ID: 6996c2b95698665ec40e0148. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. ### RAG QA Accuracy: 62.45. Result ID: 6996c2b95698665ec40e0149. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen3-8B (no-think mode)","Embedding Model":"Qwen3-8B-embedding"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.