mlfoundations-dev/

Light-R1-32B_1743609585_eval_0981

Precomputed model outputs for evaluation.

	Evaluation Results




	
	Summary




	

Metric AIME24 AIME25 AMC23 MATH500 GPQADiamond LiveCodeBench

Accuracy 74.7 58.0 96.0 90.4 62.0 56.0

Average Accuracy: 74.67% ± 1.19% Number of Runs: 5

Run Accuracy Questions Solved Total Questions

1 80.00% 24 30

2 73.33% 22 30

3 73.33% 22 30

4 73.33% 22 30

5 73.33% 22 30… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-dev/Light-R1-32B_1743609585_eval_0981.