List your company

Question Answering on TriviaQA (Calibration and Uncertainty Metrics)

5.9Error Rate (ER)Verbalized
5.13210.31615.520.684May 19, 2026

Evaluation results

MethodLinks
VerbalizedLLM=Gemini 2.5 Flash2026.055.94.974.353.25.41.650.311.36.2
Seq. Post.LLM=Gemini 2.5 Flash2026.056.46.685.520.95.71.521.312.16.4
Is TrueLLM=Gemini 2.5 Flash2026.056.45.381.358.85.33.144.911.78.1
VerbalizedLLM=Gemini 2.5 Flash Lite2026.0510.86.885.344.582.446.218.810.9
Seq. Post.LLM=Gemini 2.5 Flash Lite2026.0511.94.389.123.26.32.527.218.212
Is TrueLLM=Gemini 2.5 Flash Lite2026.0511.98.988.162.58.72.657.320.513.2
VerbalizedLLM=Ministral-3-8B-Instruct-25122026.0523.115.965.857.919.167.359.442.223.3
Seq. Post.LLM=Ministral-3-8B-Instruct-25122026.0523.730.68887.423.422.758.447.126
Is TrueLLM=Ministral-3-8B-Instruct-25122026.0523.76.389.537.310.97.34334.724.3
VerbalizedLLM=Qwen3.5 9B2026.0524.620.375.68021.210.182.345.924.8
Seq. Post.LLM=Qwen3.5 9B2026.0525.116.492.340.313.2748.438.225.2
Is TrueLLM=Qwen3.5 9B2026.0525.110.988.942.213.96.149.638.925.3