AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

UNK-VQA: A Dataset and a Probe Into the Abstention Ability of Multi-Modal Large Models

Yangyang Guo, Fangkai Jiao, Zhiqi Shen, Liqiang Nie, Mohan Kankanhalli

TPAMI 2024 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 19 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add post-hint explanation of why the question is unanswerableOtter-MPT-7B · UNK-VQA (3-shot)accuracy: 12.17 -> 67.39Improved
add post-hint explanation of why the question is unanswerableOpen-Flamingo-MPT-7B · UNK-VQA (3-shot)accuracy: 54.79 -> 55.20Improved
add post-hint explanation of why the question is unanswerableOpen-Flamingo-MPT-7B · UNK-VQA (5-shot)accuracy: 54.57 -> 57.75Improved
fine-tune LLaVA-1.6 Vicuna-13B on UNK-VQALLaVA-1.6 Vicuna-13B · UNK-VQA open-endedaccuracy: 21.53 -> 52.49Improved
fine-tune LLaVA-1.6 Mistral-7B on UNK-VQALLaVA-1.6 Mistral-7B · UNK-VQA binaryaccuracy: 56.98 -> 56.35Hurt
swap Vicuna-13B LLM for stronger Mistral-7BLLaVA-1.6 · UNK-VQA open-ended (zero-shot)accuracy: 21.53 -> 49.30Improved
scale Instruct-BLIP Vicuna-7B to 13BInstruct-BLIP · UNK-VQA zero-shot BYaccuracy: 43.98 -> 54.82Improved
scale Instruct-BLIP Vicuna-7B to 13BInstruct-BLIP · UNK-VQA zero-shot MCaccuracy: 22.38 -> 18.41Hurt
use entropy-based selective function instead of classifier-basedLXMERT · UNK-VQA testopen accuracy: 19.45 -> 22.40Improved
use entropy-based selective function instead of classifier-basedUpDn · UNK-VQA testopen accuracy: 6.46 -> 5.91No clear effect
add 2-shot demonstrationsGPT-4V · UNK-VQA subset BYaccuracy: 58.0 -> 57.0No clear effect
add 2-shot demonstrationsGPT-4V · UNK-VQA subset MCaccuracy: 38.0 -> 43.0Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy