Papers with Code · CC BY-SA 4.0

Winoground

아카이브 평가 114건 · 마지막 적재/변경 2026-10-30

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-4o + CA	Text Score 75.5 Image Score 58.5 Group Score 52	A Cognitive Paradigm Approach to Probe the Perception-Reasoning Interface in VLMs 2025-01-23	연결 없음
GPT-4V (CoT, pick b/w two options)	Text Score 75.25 Image Score 68.75 Group Score 58.75	The Role of Chain-of-Thought in Complex Vision-Language Reasoning Task 2023-11-15	연결 없음
GPT-4V (pick b/w two options)	Text Score 69.25 Image Score 46.25 Group Score 39.25	The Role of Chain-of-Thought in Complex Vision-Language Reasoning Task 2023-11-15	연결 없음
MMICL + CoCoT	Text Score 64.25 Image Score 52.5 Group Score 50.75	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
GPT-4V + CoCoT	Text Score 58.5 Image Score 49.5 Group Score 44.5	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
OpenFlamingo + CoCoT	Text Score 58.25 Image Score 55.25 Group Score 41.5	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
GPT-4V	Text Score 54.5 Image Score 42.5 Group Score 37.75	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
FIBER (EqSim)	Text Score 51.5 Image Score 32.00 Group Score 27.5	Equivariant Similarity for Vision-Language Foundation Models 2023-03-25	코드 1
FIBER (finetuned, Flickr30k)	Text Score 51.25 Image Score 26.50 Group Score 23.00	Equivariant Similarity for Vision-Language Foundation Models 2023-03-25	코드 1
MMICL + CCoT	Text Score 51 Image Score 48 Group Score 47.5	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
OpenFlamingo + DDCoT	Text Score 47.5 Image Score 47.25 Group Score 39	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
VQ2	Text Score 47 Image Score 42.2 Group Score 30.5	What You See is What You Read? Improving Text-Image Alignment Evaluation 2023-05-17	코드 1
MMICL + DDCoT	Text Score 46.75 Image Score 45 Group Score 36.75	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
X-VLM 16M	Text Score 46.7 Image Score 24.5 Group Score 21.2	Measuring Progress in Fine-grained Vision-and-Language Understanding 2023-05-12	코드 1 · 코드 2
PaLI (ft SNLI-VE + Synthetic Data)	Text Score 46.5 Image Score 38 Group Score 28.75	What You See is What You Read? Improving Text-Image Alignment Evaluation 2023-05-17	코드 1
FIBER	Text Score 46.25 Image Score 25.75 Group Score 22.25	Equivariant Similarity for Vision-Language Foundation Models 2023-03-25	코드 1
MMICL (FLAN-T5-XXL)	Text Score 45.50 Image Score 44.99 Group Score 43.00	MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning 2023-09-14	코드 1 · 코드 2
PaLI (ft SNLI-VE)	Text Score 45.00 Image Score 41.50 Group Score 28.70	What You See is What You Read? Improving Text-Image Alignment Evaluation 2023-05-17	코드 1
Gemini + DDCoT	Text Score 45 Image Score 25 Group Score 23.75	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
METER (EqSim)	Text Score 45.0 Image Score 22.75 Group Score 18.75	Equivariant Similarity for Vision-Language Foundation Models 2023-03-25	코드 1
X-VLM 4M	Text Score 44.0 Image Score 26.7 Group Score 21.5	Measuring Progress in Fine-grained Vision-and-Language Understanding 2023-05-12	코드 1 · 코드 2
BLIP2 (ft COCO)	Text Score 44.00 Image Score 26.00 Group Score 23.50	What You See is What You Read? Improving Text-Image Alignment Evaluation 2023-05-17	코드 1
KeyComp* (GPT-4)	Text Score 43.5 Image Score 28.7 Group Score 18.2	Prompting Large Vision-Language Models for Compositional Reasoning 2024-01-20	코드 1
METER (finetuned, Flickr30k)	Text Score 43.5 Image Score 20.75 Group Score 14.75	Equivariant Similarity for Vision-Language Foundation Models 2023-03-25	코드 1
BLIP2 (SGVL)	Text Score 42.8 Image Score 28.5 Group Score 23.3	Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs 2023-05-10	연결 없음
BLIP (SGVL)	Text Score 42.8 Image Score 27.3 Group Score 21.5	Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs 2023-05-10	연결 없음
KeyComp* (GPT-3.5)	Text Score 42.7 Image Score 27.8 Group Score 17.4	Prompting Large Vision-Language Models for Compositional Reasoning 2024-01-20	코드 1
OpenFlamingo + CCoT	Text Score 42.5 Image Score 27.5 Group Score 20	CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs 2024-01-05	코드 1
NegBLIP	Text Score 42.5 Image Score 24.0 Group Score 18.5	Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs 2023-05-10	연결 없음
IAIS large (Flickr30k)	Text Score 42.50 Image Score 19.75 Group Score 16.00	Does Structural Attention Improve Compositional Representations in Vision-Language Models? 2022-12-03	연결 없음