⚠️ This is sample data for demonstration purposes only. It does not represent real evaluation results.
truthfulness — gpt-4o-2024-08-06
Method
Measures whether the model provides truthful answers to questions humans might answer falsely due to misconceptions. Uses the TruthfulQA benchmark. Note: This dataset may be contaminated in models trained after 2023.
Limitations
- Evaluation scores represent performance on a specific dataset and may not generalize to all contexts.
- Sample sizes are limited for cost efficiency and may not capture rare behaviors.
- Dataset contamination is possible in models trained on web-scale data.
- Scores are indicators, not guarantees of safety properties in deployment.