README.md

mlfoundations-dev/a1_code_code_contests_eval_636d

Precomputed model outputs for evaluation.

Evaluation Results

Summary

Metric	AIME24	AMC23	MATH500	MMLUPro	JEEBench	GPQADiamond	LiveCodeBench	CodeElo	CodeForces
Accuracy	15.0	49.0	71.2	27.4	34.6	29.0	29.7	10.8	15.0

AIME24

• Average Accuracy: 15.00% ± 1.27%
• Number of Runs: 10

Run	Accuracy	Questions Solved	Total Questions
1	16.67%	5	30
2	13.33%	4	30
3	20.00%	6	30
4	20.00%	6	30
5	10.00%	3	30
6	16.67%	5	30
7	13.33%	4	30
8	10.00%	3	30
9	10.00%	3	30
10	20.00%	6	30

AMC23

• Average Accuracy: 49.00% ± 1.70%
• Number of Runs: 10

Run	Accuracy	Questions Solved	Total Questions
1	45.00%	18	40
2	52.50%	21	40
3	40.00%	16	40
4	52.50%	21	40
5	60.00%	24	40
6	52.50%	21	40
7	47.50%	19	40
8	45.00%	18	40
9	50.00%	20	40
10	45.00%	18	40

MATH500

• Accuracy: 71.20%

  Accuracy	Questions Solved	Total Questions
  71.20%	356	500

MMLUPro

• Average Accuracy: 27.40% ± 0.00%
• Number of Runs: 1

Run	Accuracy	Questions Solved	Total Questions
1	27.40%	137	500

JEEBench

• Average Accuracy: 34.64% ± 0.35%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	33.79%	174.0	515
2	35.05%	180.5	515
3	35.10%	180.75	515

GPQADiamond

• Average Accuracy: 28.96% ± 0.84%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	28.79%	57	198
2	30.81%	61	198
3	27.27%	54	198

LiveCodeBench

• Average Accuracy: 29.75% ± 1.17%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	30.92%	158	511
2	30.92%	158	511
3	27.40%	140	511

CodeElo

• Average Accuracy: 10.83% ± 0.95%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	12.02%	47	391
2	11.51%	45	391
3	8.95%	35	391

CodeForces

• Average Accuracy: 15.01% ± 0.22%
• Number of Runs: 3

Run	Accuracy	Questions Solved	Total Questions
1	15.45%	70	453
2	14.79%	67	453
3	14.79%	67	453