← Natural Questions

Papers with Code · CC BY-SA 4.0

TheoremQA

아카이브 평가 19건 · 마지막 적재/변경 2026-10-30

원본 데이터셋 링크 1

평가 기록

과거 수집 기록이며 최신 순위가 아닙니다. 데이터셋 버전·split·실험 조건과 지표 방향이 검증되지 않은 결과는 직접 비교할 수 없습니다. 날짜는 논문 날짜이며 최신 평가일을 뜻하지 않습니다.

표가 화면보다 넓으면 좌우로 스크롤하여 논문과 구현 링크를 확인하세요.

원본 기록 순서 · 순위 미부여
모델	원본 지표·값	논문	구현
GPT-4 (PoT)	Accuracy 52.4	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
GPT-4 (CoT)	Accuracy 43.8	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
GPT-3.5-turbo (PoT)	Accuracy 35.6	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
DART-Math-DSMath-7B-Uniform (0-shot CoT, w/o code)	Accuracy 32.5	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
DART-Math-DSMath-7B-Prop2Diff (0-shot CoT, w/o code)	Accuracy 32.2	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
PaLM-2-unicorn (CoT)	Accuracy 31.8	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
GPT-3.5-turbo (CoT)	Accuracy 30.2	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
DART-Math-Llama3-70B-Prop2Diff (0-shot CoT, w/o code)	Accuracy 28.2	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
DART-Math-Llama3-70B-Uniform (0-shot CoT, w/o code)	Accuracy 27.4	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
Claude-v1 (PoT)	Accuracy 25.9	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
Claude-v1 (CoT)	Accuracy 24.9	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
code-davinci-002	Accuracy 23.9	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
Claude-instant (CoT)	Accuracy 23.6	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
text-davinci-003	Accuracy 22.8	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
PaLM-2-bison (CoT)	Accuracy 21.0	TheoremQA: A Theorem-driven Question Answering dataset 2023-05-21	코드 1
DART-Math-Llama3-8B-Prop2Diff (0-shot CoT, w/o code)	Accuracy 19.4	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
DART-Math-Mistral-7B-Prop2Diff (0-shot CoT, w/o code)	Accuracy 17.0	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
DART-Math-Mistral-7B-Uniform (0-shot CoT, w/o code)	Accuracy 16.4	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1
DART-Math-Llama3-8B-Uniform (0-shot CoT, w/o code)	Accuracy 15.4	DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving 2024-06-18	코드 1