# Multi-hop Question Answering on 2Wiki (pass@1) Source: [Multi-hop Question Answering on 2Wiki (pass@1)]() Collection index: [llms.txt]() Data updated: 2026-05-20T08:23:56.564Z Evaluation of model performance on the 2Wiki dataset for multi-hop question answering, measured using the pass@1 metric. 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) | 2Wiki | | Metric | pass@1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 - [pass@1](): Higher is better; unit: PERCENT. ## Recorded results | Method | pass@1 | Result date | Report paper | | --- | --- | --- | --- | | R-Search-7B + SSP | 58 | Not recorded | 68fa7c6d1e209131cea8e038 | | R-Search-7B | 56.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen3-8B + SSP | 51.5 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-14B-Instruct + SSP | 49.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-32B-Instruct + SSP | 49.2 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-14B-Instruct | 48.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-32B-Instruct | 48.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen3-8B | 48 | Not recorded | 68fa7c6d1e209131cea8e038 | | Search-R1-7B + SSP | 45.6 | Not recorded | 68fa7c6d1e209131cea8e038 | | Search-R1-7B | 45.2 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-7B-Instruct + SSP | 38.8 | Not recorded | 68fa7c6d1e209131cea8e038 | | ZeroSearch-7B + SSP | 37.2 | Not recorded | 68fa7c6d1e209131cea8e038 | | ZeroSearch-7B | 34.6 | Not recorded | 68fa7c6d1e209131cea8e038 | | LLaMA-3.1-8B + SSP | 34.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-7B-Base + SSP | 33.2 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-7B-Instruct | 32.8 | Not recorded | 68fa7c6d1e209131cea8e038 | | LLaMA-3.1-8B | 19.4 | Not recorded | 68fa7c6d1e209131cea8e038 | | Qwen2.5-7B-Base | 10.8 | Not recorded | 68fa7c6d1e209131cea8e038 | ### R-Search-7B + SSP pass@1: 58. Result ID: 6a0d6f3cfbc373592c0569d2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"true"}. This record does not assert independent reproduction. ### R-Search-7B pass@1: 56.4. Result ID: 6a0d6f3cfbc373592c0569d1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"false"}. This record does not assert independent reproduction. ### Qwen3-8B + SSP pass@1: 51.5. Result ID: 6a0d6f3cfbc373592c0569cc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Generalization Across Model Families","SSP Integration":"true"}. This record does not assert independent reproduction. ### Qwen2.5-14B-Instruct + SSP pass@1: 49.4. Result ID: 6a0d6f3cfbc373592c0569d4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Scaling to Larger Models","SSP Integration":"true"}. This record does not assert independent reproduction. ### Qwen2.5-32B-Instruct + SSP pass@1: 49.2. Result ID: 6a0d6f3cfbc373592c0569d6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Scaling to Larger Models","SSP Integration":"true"}. This record does not assert independent reproduction. ### Qwen2.5-14B-Instruct pass@1: 48.4. Result ID: 6a0d6f3cfbc373592c0569d3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Scaling to Larger Models","SSP Integration":"false"}. This record does not assert independent reproduction. ### Qwen2.5-32B-Instruct pass@1: 48.4. Result ID: 6a0d6f3cfbc373592c0569d5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Scaling to Larger Models","SSP Integration":"false"}. This record does not assert independent reproduction. ### Qwen3-8B pass@1: 48. Result ID: 6a0d6f3cfbc373592c0569cb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Generalization Across Model Families","SSP Integration":"false"}. This record does not assert independent reproduction. ### Search-R1-7B + SSP pass@1: 45.6. Result ID: 6a0d6f3cfbc373592c0569d0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"true"}. This record does not assert independent reproduction. ### Search-R1-7B pass@1: 45.2. Result ID: 6a0d6f3cfbc373592c0569cf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"false"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct + SSP pass@1: 38.8. Result ID: 6a0d6f3cfbc373592c0569c8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"From-Scratch Training on Base and Instruct Models","SSP Integration":"true"}. This record does not assert independent reproduction. ### ZeroSearch-7B + SSP pass@1: 37.2. Result ID: 6a0d6f3cfbc373592c0569ce. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"true"}. This record does not assert independent reproduction. ### ZeroSearch-7B pass@1: 34.6. Result ID: 6a0d6f3cfbc373592c0569cd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Continual Training on Search-Specialized Agents","SSP Integration":"false"}. This record does not assert independent reproduction. ### LLaMA-3.1-8B + SSP pass@1: 34.4. Result ID: 6a0d6f3cfbc373592c0569ca. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"Generalization Across Model Families","SSP Integration":"true"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Base + SSP pass@1: 33.2. Result ID: 6a0d6f3cfbc373592c0569c6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"From-Scratch Training on Base and Instruct Models","SSP Integration":"true"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct pass@1: 32.8. Result ID: 6a0d6f3cfbc373592c0569c7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"From-Scratch Training on Base and Instruct Models","SSP Integration":"false"}. This record does not assert independent reproduction. ### LLaMA-3.1-8B pass@1: 19.4. Result ID: 6a0d6f3cfbc373592c0569c9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Training Paradigm":"Generalization Across Model Families","SSP Integration":"false"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Base pass@1: 10.8. Result ID: 6a0d6f3cfbc373592c0569c5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Paradigm":"From-Scratch Training on Base and Instruct Models","SSP Integration":"false"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.