README.md

mlfoundations-dev/a1_math_hendrycks_eval_1331

Precomputed model outputs for evaluation.

Evaluation Results

Summary

Metric	AIME24	AMC23	MATH500	MMLUPro	JEEBench	GPQADiamond	LiveCodeBench	CodeElo
Accuracy	19.7	59.0	79.4	31.2	37.6	38.2	15.1	3.8

AIME24

• Average Accuracy: 19.67% ± 1.20%
• Number of Runs: 10

Run	Accuracy	Questions Solved	Total Questions
1	16.67%	5	30
2	20.00%	6	30
3	26.67%	8	30
4	13.33%	4	30
5	23.33%	7	30
6	20.00%	6	30
7	16.67%	5	30
8	23.33%	7	30
9	20.00%	6	30
10	16.67%	5	30

AMC23

• Average Accuracy: 59.00% ± 1.38%
• Number of Runs: 10

Run	Accuracy	Questions Solved	Total Questions
1	57.50%	23	40
2	65.00%	26	40
3	65.00%	26	40
4	57.50%	23	40
5	60.00%	24	40
6	62.50%	25	40
7	60.00%	24	40
8	57.50%	23	40
9	50.00%	20	40
10	55.00%	22	40

MATH500

• Accuracy: 79.40%

  Accuracy	Questions Solved	Total Questions
  79.40%	397	500

MMLUPro

• Average Accuracy: 31.20% ± 0.00%
• Number of Runs: 1

Run	Accuracy	Questions Solved	Total Questions
1	31.20%	156	500

JEEBench

• Average Accuracy: 37.59% ± 0.42%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	37.57%	193.5	515
2	38.50%	198.25	515
3	36.70%	189.0	515

GPQADiamond

• Average Accuracy: 38.22% ± 0.73%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	37.88%	75	198
2	36.87%	73	198
3	39.90%	79	198

LiveCodeBench

• Average Accuracy: 15.07% ± 1.26%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	17.42%	89	511
2	14.68%	75	511
3	13.11%	67	511

CodeElo

• Average Accuracy: 3.84% ± 0.30%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	3.84%	15	391
2	3.32%	13	391
3	4.35%	17	391