EntropyMath Leaderboard

JWL Seedv1

5 problems 6 models

K-LLM Project · 독파모 / EntropyMath 평가 차수 안내

K-LLM Project는 대한민국 과학기술정보통신부의 「독자 AI 파운데이션 모델 프로젝트」(통칭 독파모)을 뜻합니다.

Round 1·2·3은 EntropyMath의 1차·2차·3차 모델 평가를 뜻합니다. 과기정통부의 공식 단계평가 번호나 통과 여부를 뜻하지 않습니다.

프로젝트 공식 안내 공식 1차 단계평가 공식 2차 단계평가

Jump to leaderboard

Leaderboard

Benchmarks ↑

Model	Acc	Pass@3	0	1	2	3	4
API / Others
Gemini-3-Pro-Preview	60.0	80.0	2/3	3/3	3/3	0/3	1/3
GPT-5.2 (high)	33.3	60.0	3/3	1/3	1/3	0/3	0/3
K-LLM Project · 2차 평가 (Round 2)
Solar Pro 3	40.0	40.0	3/3	0/3	3/3	0/3	0/3
K-EXAONE-236B-A23B	13.3	40.0	0/3	1/3	0/3	1/3	0/3
K-EXAONE-236B-A23B	6.7	20.0	0/3	1/3	0/3	0/3	0/3
Local - KR
Kanana-2-30B-Thinking-2601	13.3	40.0	1/3	0/3	1/3	0/3	0/3

Scroll for problems · Click a cell for the solution

All correctPartialIncorrect

Model Accuracy vs Pass@3

100%

75%

50%

25%

Gemini-3-Pro-Preview

Solar Pro 3

GPT-5.2 (high)

K-EXAONE-236B-A23B

Kanana-2-30B-Thinking-2601

K-EXAONE-236B-A23B

Accuracy

Pass@3

Avg Token Usage (Per Problem)

20,777

15,583

10,388

5,194

K-EXAONE-236B-A23B

Solar Pro 3

Gemini-3-Pro-Preview

Kanana-2-30B-Thinking-2601

GPT-5.2 (high)

Avg Tokens / Problem

About this benchmark

The JWL seed evaluation, kept separate from the standard EntropyMath seed sets.

EntropyMath is an evolutionary multi-agent system and benchmark that generates high-entropy math problems designed to systematically break current LLMs.

Results are reported using Pass@3 metrics to account for generation variance. Detailed execution traces are available for transparency.

Performance Legend

Mastery (100%)

3/3

Strong (66%)

2/3

Weak (33%)

1/3

Fail (0%)

0/3