CogVLM

Reported on 3 benchmarks across 3 tasks · 2 papers

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing1 result

Visual Question Answering (VQA)onIllusionVQA
Accuracy· 2024-03-23
38.16
best: 62.99 (GPT4-Vision 4-shot)
IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models arXiv:2403.15952

Computer Vision1 result

Object LocalizationonIllusionVQA
Accuracy· 2024-03-23
28
best: 49.7 (GPT4-Vision 4-shot+CoT)
IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models arXiv:2403.15952

Reasoning1 result

Multimodal ReasoningonREBUS
Accuracy· 2024-01-11
0.9
best: 24 (GPT-4V)
REBUS: A Robust Evaluation Benchmark of Understanding Symbols arXiv:2401.05604