# Multi-hop Question Answering on LV-Eval (test) Source: [Multi-hop Question Answering on LV-Eval (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:43:48.309Z Evaluation of model performance on the test split of the LV-Eval dataset specifically for Multi-hop Question Answering tasks, reporting F1 scores. 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) | LV-Eval (test) | | Metric | F1 Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 14 of 14 | | 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. ## Recorded results | Method | F1 Score | Result date | Report paper | | --- | --- | --- | --- | | HippoRAG2 | 12.9 | Not recorded | 699e82126017c4460ffacf5f | | HELP | 12.5 | Not recorded | 699e82126017c4460ffacf5f | | GraphRAG | 11.2 | Not recorded | 699e82126017c4460ffacf5f | | LinearRAG | 10.3 | Not recorded | 699e82126017c4460ffacf5f | | GritLM-7B | 9.8 | Not recorded | 699e82126017c4460ffacf5f | | NV-Embed-v2 | 9.8 | Not recorded | 699e82126017c4460ffacf5f | | HippoRAG | 8.4 | Not recorded | 699e82126017c4460ffacf5f | | Contriever | 8.1 | Not recorded | 699e82126017c4460ffacf5f | | GTR | 7.1 | Not recorded | 699e82126017c4460ffacf5f | | GTE-Qwen2-7B | 7.1 | Not recorded | 699e82126017c4460ffacf5f | | None | 6 | Not recorded | 699e82126017c4460ffacf5f | | BM25 | 5.9 | Not recorded | 699e82126017c4460ffacf5f | | RAPTOR | 5 | Not recorded | 699e82126017c4460ffacf5f | | LightRAG | 1 | Not recorded | 699e82126017c4460ffacf5f | ### HippoRAG2 F1 Score: 12.9. Result ID: 699ebc2c5698665ec48b6929. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### HELP F1 Score: 12.5. Result ID: 699ebc2c5698665ec48b692b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### GraphRAG F1 Score: 11.2. Result ID: 699ebc2c5698665ec48b6926. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### LinearRAG F1 Score: 10.3. Result ID: 699ebc2c5698665ec48b692a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct","Local Reproduction":true}. This record does not assert independent reproduction. ### GritLM-7B F1 Score: 9.8. Result ID: 699ebc2c5698665ec48b6923. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### NV-Embed-v2 F1 Score: 9.8. Result ID: 699ebc2c5698665ec48b6924. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### HippoRAG F1 Score: 8.4. Result ID: 699ebc2c5698665ec48b6928. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### Contriever F1 Score: 8.1. Result ID: 699ebc2c5698665ec48b691f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### GTR F1 Score: 7.1. Result ID: 699ebc2c5698665ec48b6921. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### GTE-Qwen2-7B F1 Score: 7.1. Result ID: 699ebc2c5698665ec48b6922. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### None F1 Score: 6. Result ID: 699ebc2c5698665ec48b691e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### BM25 F1 Score: 5.9. Result ID: 699ebc2c5698665ec48b6920. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### RAPTOR F1 Score: 5. Result ID: 699ebc2c5698665ec48b6925. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. ### LightRAG F1 Score: 1. Result ID: 699ebc2c5698665ec48b6927. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-3.3-70B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.