# Truthful Question Answering on TruthfulQA (TruthRate and InfoRate) Source: [Truthful Question Answering on TruthfulQA (TruthRate and InfoRate)]() Collection index: [llms.txt]() Data updated: 2026-06-16T08:51:42.627Z Evaluation of model performance on the TruthfulQA dataset, focusing on the TruthRate and InfoRate metrics. 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 | Truthful Question Answering | | Dataset (including any stated split) | TruthfulQA | | Metric | TruthRate | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [TruthRate](): Higher is better; unit: PERCENT. - [InfoRate](): Higher is better; unit: PERCENT. ## Recorded results | Method | TruthRate | Result date | Report paper | | --- | --- | --- | --- | | SEA | 83 | Not recorded | 68fa7c011e209131cea8bd43 | | BoN64 | 76 | Not recorded | 68fa7c011e209131cea8bd43 | | SFT | 73 | Not recorded | 68fa7c011e209131cea8bd43 | | ITI | 57.64 | Not recorded | 6a30d8a4d33dfb23e783ccdd | | TruthProbeBase LLM | 56.91 | Not recorded | 6a30d8a4d33dfb23e783ccdd | | LLaMA2-7B-Chat | 56.4 | Not recorded | 6a30d8a4d33dfb23e783ccdd | | TruthProbeFT LLM | 55.38 | Not recorded | 6a30d8a4d33dfb23e783ccdd | | L3-CAT | 43 | Not recorded | 6a02b42bdbc836b87d824771 | | LPA (ours) | 38.1 | Not recorded | 6a02b42bdbc836b87d824771 | | Base (8B Instruct) | 37 | Not recorded | 6a02b42bdbc836b87d824771 | | L2-CAT | 35.3 | Not recorded | 6a02b42bdbc836b87d824771 | | LPA-overfit (ours) | 33.8 | Not recorded | 6a02b42bdbc836b87d824771 | | L3-LAT | 33.2 | Not recorded | 6a02b42bdbc836b87d824771 | | LPA (ours) | 31.5 | Not recorded | 6a02b42bdbc836b87d824771 | | Base (7B chat-hf) | 30.7 | Not recorded | 6a02b42bdbc836b87d824771 | | LPA-overfit (ours) | 30.6 | Not recorded | 6a02b42bdbc836b87d824771 | | L2-LAT | 28.6 | Not recorded | 6a02b42bdbc836b87d824771 | ### SEA TruthRate: 83. Result ID: 69b91bae47f34ce924a1f1f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-13B-Chat"}. This record does not assert independent reproduction. ### BoN64 TruthRate: 76. Result ID: 69b91bae47f34ce924a1f1f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-13B-Chat"}. This record does not assert independent reproduction. ### SFT TruthRate: 73. Result ID: 69b91bae47f34ce924a1f1f6. [Reporting paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-13B-Chat"}. This record does not assert independent reproduction. ### ITI TruthRate: 57.64. Result ID: 6a310e8bfbc373592c2d7910. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B-Chat"}. This record does not assert independent reproduction. ### TruthProbeBase LLM TruthRate: 56.91. Result ID: 6a310e8bfbc373592c2d7911. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B-Chat"}. This record does not assert independent reproduction. ### LLaMA2-7B-Chat TruthRate: 56.4. Result ID: 6a310e8bfbc373592c2d790f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Mode":"Vanilla"}. This record does not assert independent reproduction. ### TruthProbeFT LLM TruthRate: 55.38. Result ID: 6a310e8bfbc373592c2d7912. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"LLaMA2-7B-Chat"}. This record does not assert independent reproduction. ### L3-CAT TruthRate: 43. Result ID: 6a02e7c247f34ce924e5b8b8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3"}. This record does not assert independent reproduction. ### LPA (ours) TruthRate: 38.1. Result ID: 6a02e7c247f34ce924e5b8b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3"}. This record does not assert independent reproduction. ### Base (8B Instruct) TruthRate: 37. Result ID: 6a02e7c247f34ce924e5b8b6. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama 3"}. This record does not assert independent reproduction. ### L2-CAT TruthRate: 35.3. Result ID: 6a02e7c247f34ce924e5b8bd. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 2"}. This record does not assert independent reproduction. ### LPA-overfit (ours) TruthRate: 33.8. Result ID: 6a02e7c247f34ce924e5b8ba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3"}. This record does not assert independent reproduction. ### L3-LAT TruthRate: 33.2. Result ID: 6a02e7c247f34ce924e5b8b7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3"}. This record does not assert independent reproduction. ### LPA (ours) TruthRate: 31.5. Result ID: 6a02e7c247f34ce924e5b8be. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 2"}. This record does not assert independent reproduction. ### Base (7B chat-hf) TruthRate: 30.7. Result ID: 6a02e7c247f34ce924e5b8bb. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama 2"}. This record does not assert independent reproduction. ### LPA-overfit (ours) TruthRate: 30.6. Result ID: 6a02e7c247f34ce924e5b8bf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 2"}. This record does not assert independent reproduction. ### L2-LAT TruthRate: 28.6. Result ID: 6a02e7c247f34ce924e5b8bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 2"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.