mlfoundations-dev/

gpt-4.1-2025-04-14_eval_68b3

Precomputed model outputs for evaluation.

	Evaluation Results




	
	Summary




	

Metric MATH500 GPQADiamond CodeElo MMLUPro JEEBench LiveCodeBenchv5 LiveCodeBench AIME25 AIME24 AMC23 HLE

Accuracy 83.6 34.5 31.1 30.8 78.3 46.6 65.6 33.0 50.7 83.2 7.2

Accuracy: 83.6%

Accuracy Questions Solved Total Questions

83.6% 418 500

Average Accuracy: 34.5% ± 0.4%… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-dev/gpt-4.1-2025-04-14_eval_68b3.