# Question Answering on NQ-Open In-Domain (test) Source: [Question Answering on NQ-Open In-Domain (test)]() Collection index: [llms.txt]() Data updated: 2026-04-14T08:50:59.058Z Evaluation of Question Answering performance on the NQ-Open In-Domain test set, reporting Precision and Truthfulness 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 | Question Answering | | Dataset (including any stated split) | NQ-Open In-Domain (test) | | Metric | Precision | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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 - [Precision](): Higher is better; unit: PERCENT. - [Truthfulness](): Higher is better; unit: PERCENT. ## Recorded results | Method | Precision | Result date | Report paper | | --- | --- | --- | --- | | FAITH | 58.13 | Not recorded | 69ddca337d0172140ec79b27 | | FAITH\_sft+ppo | 57.99 | Not recorded | 69ddca337d0172140ec79b27 | | ICL-CoT | 57.34 | Not recorded | 69ddca337d0172140ec79b27 | | InFact | 56.93 | Not recorded | 69ddca337d0172140ec79b27 | | Context-DPO | 56.82 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign\_rag | 56.78 | Not recorded | 69ddca337d0172140ec79b27 | | DTA | 56.72 | Not recorded | 69ddca337d0172140ec79b27 | | FAITH\_sft | 56.72 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign | 56.6 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign\_sft | 56.02 | Not recorded | 69ddca337d0172140ec79b27 | | ICL-CoT | 54.47 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign\_rag | 54.47 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign | 54.17 | Not recorded | 69ddca337d0172140ec79b27 | | SFT\_rag | 51.03 | Not recorded | 69ddca337d0172140ec79b27 | | RL-DPO | 50.95 | Not recorded | 69ddca337d0172140ec79b27 | | SFT | 50.82 | Not recorded | 69ddca337d0172140ec79b27 | | RL-DPO | 49.65 | Not recorded | 69ddca337d0172140ec79b27 | | FAITH | 48.05 | Not recorded | 69ddca337d0172140ec79b27 | | Context-DPO | 47.21 | Not recorded | 69ddca337d0172140ec79b27 | | InFact | 47.01 | Not recorded | 69ddca337d0172140ec79b27 | | FAITH\_sft+ppo | 46.6 | Not recorded | 69ddca337d0172140ec79b27 | | FAITH\_sft | 46.16 | Not recorded | 69ddca337d0172140ec79b27 | | UAlign\_sft | 43.32 | Not recorded | 69ddca337d0172140ec79b27 | | SFT\_rag | 42.64 | Not recorded | 69ddca337d0172140ec79b27 | | SFT | 41.41 | Not recorded | 69ddca337d0172140ec79b27 | | DTA | 41.01 | Not recorded | 69ddca337d0172140ec79b27 | ### FAITH Precision: 58.13. Result ID: 69ddfecc47f34ce924c57d92. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### FAITH\_sft+ppo Precision: 57.99. Result ID: 69ddfecc47f34ce924c57d93. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### ICL-CoT Precision: 57.34. Result ID: 69ddfecc47f34ce924c57d88. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### InFact Precision: 56.93. Result ID: 69ddfecc47f34ce924c57d91. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### Context-DPO Precision: 56.82. Result ID: 69ddfecc47f34ce924c57d90. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### UAlign\_rag Precision: 56.78. Result ID: 69ddfecc47f34ce924c57d8f. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### DTA Precision: 56.72. Result ID: 69ddfecc47f34ce924c57d8c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### FAITH\_sft Precision: 56.72. Result ID: 69ddfecc47f34ce924c57d94. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### UAlign Precision: 56.6. Result ID: 69ddfecc47f34ce924c57d8d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### UAlign\_sft Precision: 56.02. Result ID: 69ddfecc47f34ce924c57d8e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### ICL-CoT Precision: 54.47. Result ID: 69ddfecc47f34ce924c57d95. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### UAlign\_rag Precision: 54.47. Result ID: 69ddfecc47f34ce924c57d9c. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### UAlign Precision: 54.17. Result ID: 69ddfecc47f34ce924c57d9a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### SFT\_rag Precision: 51.03. Result ID: 69ddfecc47f34ce924c57d97. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### RL-DPO Precision: 50.95. Result ID: 69ddfecc47f34ce924c57d98. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### SFT Precision: 50.82. Result ID: 69ddfecc47f34ce924c57d96. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### RL-DPO Precision: 49.65. Result ID: 69ddfecc47f34ce924c57d8b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### FAITH Precision: 48.05. Result ID: 69ddfecc47f34ce924c57d9f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### Context-DPO Precision: 47.21. Result ID: 69ddfecc47f34ce924c57d9d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### InFact Precision: 47.01. Result ID: 69ddfecc47f34ce924c57d9e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### FAITH\_sft+ppo Precision: 46.6. Result ID: 69ddfecc47f34ce924c57da0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### FAITH\_sft Precision: 46.16. Result ID: 69ddfecc47f34ce924c57da1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### UAlign\_sft Precision: 43.32. Result ID: 69ddfecc47f34ce924c57d9b. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. ### SFT\_rag Precision: 42.64. Result ID: 69ddfecc47f34ce924c57d8a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### SFT Precision: 41.41. Result ID: 69ddfecc47f34ce924c57d89. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama3-8B"}. This record does not assert independent reproduction. ### DTA Precision: 41.01. Result ID: 69ddfecc47f34ce924c57d99. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-v0.1"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.