Papers with Code · CC BY-SA 4.0

MM-Vet

아카이브 평가 231건 · 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
TACO (Qwen2-7B / SigLIP)	GPT-4 score 50.9	TACO: Learning Multi-modal Action Models with Synthetic Chains-of-Thought-and-Action 2024-12-07	코드 1
Qwen-VL-Chat (+ SFT (GPT-4V in VLFeedback))	GPT-4 score 50.7	VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment 2024-10-12	연결 없음
POINTS-9B	GPT-4 score 50.0	POINTS: Improving Your Vision-language Model with Affordable Strategies 2024-09-07	연결 없음
VILA^2-8B	GPT-4 score 50.0	VILA$^2$: VILA Augmented VILA 2024-07-24	연결 없음
Janus-Pro-7B	GPT-4 score 50.0	Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling 2025-01-29	코드 1
Silkie	GPT-4 score 49.9 Params 7B	Silkie: Preference Distillation for Large Visual Language Models 2023-12-17	연결 없음
Silkie (Qwen-VL-Chat + DPO w/ VLFeedback)	GPT-4 score 49.9	VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment 2024-10-12	연결 없음
Qwen2-VL-2B	GPT-4 score 49.5	Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution 2024-09-18	코드 1 · 코드 2 · 코드 3 · 코드 4 · 코드 5 · 코드 6 · 코드 7 · 코드 8
FlashSloth-HD	GPT-4 score 49.0 Params 3.2B	FlashSloth: Lightning Multimodal Large Language Models via Embedded Visual Compression 2024-12-05	코드 1
InternVL 1.2	GPT-4 score 48.9 Params 40B	How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites 2024-04-25	코드 1
SEA-PRIME (Vicuna-13B)	GPT-4 score 48.8	SEA: Supervised Embedding Alignment for Token-Level Visual-Textual Integration in MLLMs 2024-08-21	연결 없음
InternVL2.5-1B	GPT-4 score 48.8 Params 1B	Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling 2024-12-06	코드 1
MM1-30B-Chat	GPT-4 score 48.7	MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training 2024-03-14	연결 없음
SETOKIM (13B)	GPT-4 score 48.7 Params 13B	Towards Semantic Equivalence of Tokenization in Multimodal LLM 2024-06-07	연결 없음
Emu2-Chat	GPT-4 score 48.5 Params 37B	Generative Multimodal Models are In-Context Learners 2023-12-20	코드 1
MG-LLaVA(34B)	GPT-4 score 48.5	MG-LLaVA: Towards Multi-Granularity Visual Instruction Tuning 2024-06-25	코드 1
SPHINX-Plus	GPT-4 score 47.9	SPHINX-X: Scaling Data and Parameters for a Family of Multi-modal Large Language Models 2024-02-08	코드 1
ConvLLaVA	GPT-4 score 45.9 Params 7B	ConvLLaVA: Hierarchical Backbones as Visual Encoder for Large Multimodal Models 2024-05-24	코드 1
VILA-13B	GPT-4 score 45.7	VILA: On Pre-training for Visual Language Models 2023-12-12	코드 1 · 코드 2 · 코드 3
TACO (LLaMA3-8B / SigLIP)	GPT-4 score 45.7	TACO: Learning Multi-modal Action Models with Synthetic Chains-of-Thought-and-Action 2024-12-07	코드 1
HPT 1.5 Edge	GPT-4 score 45.3	연결 없음 논문 날짜 미기재	연결 없음
TACO (LLaMA3-8B / CLIP)	GPT-4 score 45.2	TACO: Learning Multi-modal Action Models with Synthetic Chains-of-Thought-and-Action 2024-12-07	코드 1
LLaVA-v1.6 (7B, w/ STIC)	GPT-4 score 45.0 Params 7B	Enhancing Large Vision Language Models with Self-Training on Image Comprehension 2024-05-30	코드 1
H2OVL-Mississippi-2B	GPT-4 score 44.7	H2OVL-Mississippi Vision Language Models Technical Report 2024-10-17	연결 없음
PIIP-LLaVA (Vicuna-7B, ConvNeXt-L, CLIP-L )	GPT-4 score 44.7 Params 7B	Parameter-Inverted Image Pyramid Networks for Visual Perception and Multimodal Understanding 2025-01-14	코드 1
MM-ReAct-GPT-4	GPT-4 score 44.6±0.2	MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action 2023-03-20	코드 1
Imp-4B	GPT-4 score 44.6 Params 4B	Imp: Highly Capable Large Multimodal Models for Mobile Devices 2024-05-20	코드 1
LLaVA-Next-Mistral-7b (+ DPO w/ VLFeedback)	GPT-4 score 44.2	VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment 2024-10-12	연결 없음
MGM-7B+RP	GPT-4 score 44.1	Img-Diff: Contrastive Data Synthesis for Multimodal Large Language Models 2024-08-08	연결 없음
LLaVA-Next-Vicuna-7b (+ DPO w/ VLFeedback)	GPT-4 score 44.1	VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment 2024-10-12	연결 없음