# Multi-hop Question Answering on Dragonball DragBalance (test) Source: [Multi-hop Question Answering on Dragonball DragBalance (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:01:13.887Z Evaluation of model performance on the test split of the Dragonball DragBalance dataset for Multi-hop Question Answering. 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) | Dragonball DragBalance (test) | | Metric | Recall | | 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 - [Recall](): Higher is better; unit: PERCENT. - [EIR](): Lower is better; unit: PERCENT. - [Comp.](): Higher is better; unit: PERCENT. - [Hall.](): Lower is better; unit: PERCENT. - [Irr.](): Lower is better; unit: PERCENT. ## Recorded results | Method | Recall | Result date | Report paper | | --- | --- | --- | --- | | Qwen3-32B | 100 | Not recorded | 697846c1a2b28e761d87c694 | | Gemini-2.5-Flash | 100 | Not recorded | 697846c1a2b28e761d87c694 | | Gemini-2.5-Pro | 100 | Not recorded | 697846c1a2b28e761d87c694 | | FABLE(docs) | 85.8 | Not recorded | 697846c1a2b28e761d87c694 | | FABLE(llm-docs) | 85.55 | Not recorded | 697846c1a2b28e761d87c694 | | FABLE(nodes) | 84.55 | Not recorded | 697846c1a2b28e761d87c694 | | FABLE(llm-nodes) | 77.3 | Not recorded | 697846c1a2b28e761d87c694 | | BM25 | 66.14 | Not recorded | 697846c1a2b28e761d87c694 | | BGE-M3 | 64.02 | Not recorded | 697846c1a2b28e761d87c694 | | HippoRAG2 | 39.16 | Not recorded | 697846c1a2b28e761d87c694 | | LongRefiner | 24.72 | Not recorded | 697846c1a2b28e761d87c694 | ### Qwen3-32B Recall: 100. Result ID: 69966dba5698665ec4b49475. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"category":"LLM-in-context","mode":"Full-context Inference"}. This record does not assert independent reproduction. ### Gemini-2.5-Flash Recall: 100. Result ID: 69966dba5698665ec4b49476. [Reporting paper]() Recorded evaluation configuration: {"category":"LLM-in-context","mode":"Full-context Inference"}. This record does not assert independent reproduction. ### Gemini-2.5-Pro Recall: 100. Result ID: 69966dba5698665ec4b49477. [Reporting paper]() Recorded evaluation configuration: {"category":"LLM-in-context","mode":"Full-context Inference"}. This record does not assert independent reproduction. ### FABLE(docs) Recall: 85.8. Result ID: 69966dba5698665ec4b4947a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"selection\_strategy":"LLM-guided","bi\_path\_fusion":true,"retrieval\_unit":"documents"}. This record does not assert independent reproduction. ### FABLE(llm-docs) Recall: 85.55. Result ID: 69966dba5698665ec4b49478. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"selection\_strategy":"LLM-guided only","bi\_path\_fusion":false,"retrieval\_unit":"documents"}. This record does not assert independent reproduction. ### FABLE(nodes) Recall: 84.55. Result ID: 69966dba5698665ec4b4947b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"selection\_strategy":"LLM-guided","bi\_path\_fusion":true,"retrieval\_unit":"nodes"}. This record does not assert independent reproduction. ### FABLE(llm-nodes) Recall: 77.3. Result ID: 69966dba5698665ec4b49479. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"selection\_strategy":"LLM-guided only","bi\_path\_fusion":false,"retrieval\_unit":"nodes"}. This record does not assert independent reproduction. ### BM25 Recall: 66.14. Result ID: 69966dba5698665ec4b49471. [Reporting paper]() Recorded evaluation configuration: {"category":"Traditional RAG"}. This record does not assert independent reproduction. ### BGE-M3 Recall: 64.02. Result ID: 69966dba5698665ec4b49472. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"category":"Traditional RAG"}. This record does not assert independent reproduction. ### HippoRAG2 Recall: 39.16. Result ID: 69966dba5698665ec4b49474. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"category":"Structure-Enhanced RAG"}. This record does not assert independent reproduction. ### LongRefiner Recall: 24.72. Result ID: 69966dba5698665ec4b49473. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"category":"Structure-Enhanced RAG"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.