# Multihop Question Answering on 2WikiMultihopQA (EM, ASR) Source: [Multihop Question Answering on 2WikiMultihopQA (EM, ASR)]() Collection index: [llms.txt]() Data updated: 2026-07-02T08:40:29.052Z Evaluation of model performance on the 2WikiMultihopQA dataset for multihop question answering, reporting Exact Match (EM) and Answer Success Rate (ASR). 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 | Multihop Question Answering | | Dataset (including any stated split) | 2WikiMultihopQA | | Metric | EM | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 36 of 36 | | 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 - [EM](): Higher is better; unit: PERCENT. - [ASR](): Higher is better; unit: PERCENT. ## Recorded results | Method | EM | Result date | Report paper | | --- | --- | --- | --- | | Clean (No Attack) | 60 | Not recorded | 6a45f087dc71372d7795a25c | | Clean (No Attack) | 58 | Not recorded | 6a45f087dc71372d7795a25c | | PARADOX | 57 | Not recorded | 6a45f087dc71372d7795a25c | | Clean (No Attack) | 57 | Not recorded | 6a45f087dc71372d7795a25c | | Ignore | 56 | Not recorded | 6a45f087dc71372d7795a25c | | PARADOX | 55 | Not recorded | 6a45f087dc71372d7795a25c | | Clean (No Attack) | 54 | Not recorded | 6a45f087dc71372d7795a25c | | Fake Completion | 54 | Not recorded | 6a45f087dc71372d7795a25c | | Fake Completion | 54 | Not recorded | 6a45f087dc71372d7795a25c | | Combined | 53 | Not recorded | 6a45f087dc71372d7795a25c | | Combined | 52 | Not recorded | 6a45f087dc71372d7795a25c | | Topic | 52 | Not recorded | 6a45f087dc71372d7795a25c | | PARADOX | 52 | Not recorded | 6a45f087dc71372d7795a25c | | PARADOX | 51 | Not recorded | 6a45f087dc71372d7795a25c | | Naive | 51 | Not recorded | 6a45f087dc71372d7795a25c | | Topic | 51 | Not recorded | 6a45f087dc71372d7795a25c | | Fake Completion | 50 | Not recorded | 6a45f087dc71372d7795a25c | | Naive | 50 | Not recorded | 6a45f087dc71372d7795a25c | | Ignore | 46 | Not recorded | 6a45f087dc71372d7795a25c | | PoisonedRAG | 44 | Not recorded | 6a45f087dc71372d7795a25c | | Naive | 42 | Not recorded | 6a45f087dc71372d7795a25c | | Naive | 41 | Not recorded | 6a45f087dc71372d7795a25c | | PoisonedRAG | 37 | Not recorded | 6a45f087dc71372d7795a25c | | PoisonedRAG | 36 | Not recorded | 6a45f087dc71372d7795a25c | | Ignore | 35 | Not recorded | 6a45f087dc71372d7795a25c | | PoisonedRAG | 34 | Not recorded | 6a45f087dc71372d7795a25c | | Combined | 32 | Not recorded | 6a45f087dc71372d7795a25c | | KidnapRAG | 31 | Not recorded | 6a45f087dc71372d7795a25c | | Combined | 26 | Not recorded | 6a45f087dc71372d7795a25c | | KidnapRAG | 25 | Not recorded | 6a45f087dc71372d7795a25c | | Fake Completion | 23 | Not recorded | 6a45f087dc71372d7795a25c | | Ignore | 21 | Not recorded | 6a45f087dc71372d7795a25c | | Topic | 16 | Not recorded | 6a45f087dc71372d7795a25c | | Topic | 16 | Not recorded | 6a45f087dc71372d7795a25c | | KidnapRAG | 10 | Not recorded | 6a45f087dc71372d7795a25c | | KidnapRAG | 9 | Not recorded | 6a45f087dc71372d7795a25c | ### Clean (No Attack) EM: 60. Result ID: 6a4623fb60b88bb339854074. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### Clean (No Attack) EM: 58. Result ID: 6a4623fb60b88bb33985408f. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### PARADOX EM: 57. Result ID: 6a4623fb60b88bb339854084. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Clean (No Attack) EM: 57. Result ID: 6a4623fb60b88bb339854086. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Ignore EM: 56. Result ID: 6a4623fb60b88bb339854088. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### PARADOX EM: 55. Result ID: 6a4623fb60b88bb33985408d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Clean (No Attack) EM: 54. Result ID: 6a4623fb60b88bb33985407d. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Fake Completion EM: 54. Result ID: 6a4623fb60b88bb339854089. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Fake Completion EM: 54. Result ID: 6a4623fb60b88bb339854092. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### Combined EM: 53. Result ID: 6a4623fb60b88bb339854093. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### Combined EM: 52. Result ID: 6a4623fb60b88bb33985408a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Topic EM: 52. Result ID: 6a4623fb60b88bb339854094. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### PARADOX EM: 52. Result ID: 6a4623fb60b88bb339854096. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### PARADOX EM: 51. Result ID: 6a4623fb60b88bb33985407b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### Naive EM: 51. Result ID: 6a4623fb60b88bb339854087. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Topic EM: 51. Result ID: 6a4623fb60b88bb33985408b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Fake Completion EM: 50. Result ID: 6a4623fb60b88bb339854080. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Naive EM: 50. Result ID: 6a4623fb60b88bb339854090. [Reporting paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### Ignore EM: 46. Result ID: 6a4623fb60b88bb339854091. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### PoisonedRAG EM: 44. Result ID: 6a4623fb60b88bb339854083. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Naive EM: 42. Result ID: 6a4623fb60b88bb33985407e. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Naive EM: 41. Result ID: 6a4623fb60b88bb339854075. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### PoisonedRAG EM: 37. Result ID: 6a4623fb60b88bb33985408c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### PoisonedRAG EM: 36. Result ID: 6a4623fb60b88bb339854095. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### Ignore EM: 35. Result ID: 6a4623fb60b88bb339854076. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### PoisonedRAG EM: 34. Result ID: 6a4623fb60b88bb33985407a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### Combined EM: 32. Result ID: 6a4623fb60b88bb339854081. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### KidnapRAG EM: 31. Result ID: 6a4623fb60b88bb339854097. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"DeepSeek-32B"}. This record does not assert independent reproduction. ### Combined EM: 26. Result ID: 6a4623fb60b88bb339854078. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### KidnapRAG EM: 25. Result ID: 6a4623fb60b88bb33985408e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"WebThinker","LLM Backbone":"QwQ-32B"}. This record does not assert independent reproduction. ### Fake Completion EM: 23. Result ID: 6a4623fb60b88bb339854077. [Reporting paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### Ignore EM: 21. Result ID: 6a4623fb60b88bb33985407f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### Topic EM: 16. Result ID: 6a4623fb60b88bb339854079. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. ### Topic EM: 16. Result ID: 6a4623fb60b88bb339854082. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### KidnapRAG EM: 10. Result ID: 6a4623fb60b88bb339854085. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Llama-3.3-70B-Inst"}. This record does not assert independent reproduction. ### KidnapRAG EM: 9. Result ID: 6a4623fb60b88bb33985407c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Framework":"ReAct","LLM Backbone":"Qwen2.5-32B-Inst"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.