# Fact-Checking on PubHealth Source: [Fact-Checking on PubHealth]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:37:33.622Z Evaluation of model performance on the PubHealth dataset for Fact-Checking tasks, measured using Balanced Accuracy. 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 | Fact-Checking | | Dataset (including any stated split) | PubHealth | | Metric | Balanced Accuracy | | 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 - [Balanced Accuracy](): Higher is better; unit: PERCENT. - [Precision](): Higher is better; unit: PERCENT. - [Recall](): Higher is better; unit: PERCENT. - [F1 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Balanced Accuracy | Result date | Report paper | | --- | --- | --- | --- | | KG-CRAFTL3.3 | 78.66 | Not recorded | 69799817a2b28e761d87cc3f | | Llama3.3 70B | 76.9 | Not recorded | 68fa78d41e209131cea7aa12 | | Llama3.3 70B | 76.9 | Not recorded | 68fa78d41e209131cea7aa12 | | Qwen2.5 72B | 76.7 | Not recorded | 68fa78d41e209131cea7aa12 | | Qwen2.5 72B | 76.7 | Not recorded | 68fa78d41e209131cea7aa12 | | OpenAI o1 | 74.8 | Not recorded | 68fa78d41e209131cea7aa12 | | Claude 3.5-Sonnet | 73.8 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheck-Qwen2.5 72B | 73.6 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheck | 73.6 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheckL3.3 | 73.6 | Not recorded | 69799817a2b28e761d87cc3f | | GPT-4 | 73.2 | Not recorded | 68fa78d41e209131cea7aa12 | | DeepSeek-V3 671B | 72.9 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheck-Llama3.3 70B | 71.7 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheck | 71.7 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphCheckQwen 72B | 71.7 | Not recorded | 69799817a2b28e761d87cc3f | | Llama3 8B | 70.7 | Not recorded | 68fa78d41e209131cea7aa12 | | GPT-4o | 67 | Not recorded | 68fa78d41e209131cea7aa12 | | MiniCheck | 66.3 | Not recorded | 68fa78d41e209131cea7aa12 | | MiniCheck | 66.3 | Not recorded | 68fa78d41e209131cea7aa12 | | AlignScore | 64.4 | Not recorded | 68fa78d41e209131cea7aa12 | | AlignScore | 64.4 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphEval | 63.7 | Not recorded | 68fa78d41e209131cea7aa12 | | GraphEval | 63.7 | Not recorded | 68fa78d41e209131cea7aa12 | | ACUEval | 62.9 | Not recorded | 68fa78d41e209131cea7aa12 | | KG-CRAFTC3.5 | 61.02 | Not recorded | 69799817a2b28e761d87cc3f | | Qwen2.5 7B | 59.1 | Not recorded | 68fa78d41e209131cea7aa12 | ### KG-CRAFTL3.3 Balanced Accuracy: 78.66. Result ID: 6996693b5698665ec4afce49. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"L3.3"}. This record does not assert independent reproduction. ### Llama3.3 70B Balanced Accuracy: 76.9. Result ID: 6995a4a93badfaffe6a208ae. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama3.3 70B"}. This record does not assert independent reproduction. ### Llama3.3 70B Balanced Accuracy: 76.9. Result ID: 6995a4a93badfaffe6a20928. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5 72B Balanced Accuracy: 76.7. Result ID: 6995a4a93badfaffe6a208ad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5 72B"}. This record does not assert independent reproduction. ### Qwen2.5 72B Balanced Accuracy: 76.7. Result ID: 6995a4a93badfaffe6a20929. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### OpenAI o1 Balanced Accuracy: 74.8. Result ID: 6995a4a93badfaffe6a20923. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Scale":"\>300B"}. This record does not assert independent reproduction. ### Claude 3.5-Sonnet Balanced Accuracy: 73.8. Result ID: 6995a4a93badfaffe6a20924. [Reporting paper]() Recorded evaluation configuration: {"Scale":"\>300B"}. This record does not assert independent reproduction. ### GraphCheck-Qwen2.5 72B Balanced Accuracy: 73.6. Result ID: 6995a4a93badfaffe6a208b2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5 72B"}. This record does not assert independent reproduction. ### GraphCheck Balanced Accuracy: 73.6. Result ID: 6995a4a93badfaffe6a2092e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama3.3 70B"}. This record does not assert independent reproduction. ### GraphCheckL3.3 Balanced Accuracy: 73.6. Result ID: 6996693b5698665ec4afce46. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"L3.3"}. This record does not assert independent reproduction. ### GPT-4 Balanced Accuracy: 73.2. Result ID: 6995a4a93badfaffe6a20921. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Scale":"\>300B"}. This record does not assert independent reproduction. ### DeepSeek-V3 671B Balanced Accuracy: 72.9. Result ID: 6995a4a93badfaffe6a20925. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Scale":"\>300B"}. This record does not assert independent reproduction. ### GraphCheck-Llama3.3 70B Balanced Accuracy: 71.7. Result ID: 6995a4a93badfaffe6a208b3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama3.3 70B"}. This record does not assert independent reproduction. ### GraphCheck Balanced Accuracy: 71.7. Result ID: 6995a4a93badfaffe6a2092f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen 72B"}. This record does not assert independent reproduction. ### GraphCheckQwen 72B Balanced Accuracy: 71.7. Result ID: 6996693b5698665ec4afce47. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen 72B"}. This record does not assert independent reproduction. ### Llama3 8B Balanced Accuracy: 70.7. Result ID: 6995a4a93badfaffe6a20926. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GPT-4o Balanced Accuracy: 67. Result ID: 6995a4a93badfaffe6a20922. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Scale":"\>300B"}. This record does not assert independent reproduction. ### MiniCheck Balanced Accuracy: 66.3. Result ID: 6995a4a93badfaffe6a208b0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MiniCheck Balanced Accuracy: 66.3. Result ID: 6995a4a93badfaffe6a2092c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AlignScore Balanced Accuracy: 64.4. Result ID: 6995a4a93badfaffe6a208af. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AlignScore Balanced Accuracy: 64.4. Result ID: 6995a4a93badfaffe6a2092a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GraphEval Balanced Accuracy: 63.7. Result ID: 6995a4a93badfaffe6a208b1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### GraphEval Balanced Accuracy: 63.7. Result ID: 6995a4a93badfaffe6a2092d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ACUEval Balanced Accuracy: 62.9. Result ID: 6995a4a93badfaffe6a2092b. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### KG-CRAFTC3.5 Balanced Accuracy: 61.02. Result ID: 6996693b5698665ec4afce48. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"C3.5"}. This record does not assert independent reproduction. ### Qwen2.5 7B Balanced Accuracy: 59.1. Result ID: 6995a4a93badfaffe6a20927. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.