AI Paper Atlas

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

Choices Speak Louder than Questions

Gyeongje Cho, Yeonkyoung So, Jaejin Lee

ICLR 2026 · Vision-language and multimodal › Visual question answering · approach: Vision-language model VQA evaluation · 5 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
replace log-likelihood scoring with NPSQ under simple adversarial choiceLlama-3.1-8B-Instruct · HellaSwagpredictions switched to adversarial: 93.19% -> under 0.17%Improved
replace length-normalized scoring with NPSQ under extended adversarial choiceLlama-3.1-8B-Instruct · ARC-Challengeaccuracy drop: 18.17% drop -> below 0.05% differenceImproved
replace acc with NPSQ under instructional adversarial choiceLlama-3.1-8B-Instruct · MMLUpredictions switched: 27.47% -> 10.13%Improved
replace acc with NPSQ under neutral adversarial choiceLlama-3.1-8B-Instruct · MMLUaccuracy change: -8.60% (acc) -> +3.31% (NPSQ)Improved
use NPSQ instead of acc in cloze formatLlama-3.1-8B-Instruct · MMLUreported accuracy: higher score than acc and acc_normImproved
use NPSQ instead of acc in symbols and hybrid formatsLlama-3.1-8B-Instruct · MMLUreported accuracy: slightly lower scoresHurt
use NPSQ for ranking modelsvarious LLMs · MCQA benchmarksmodel ranking: rankings reorder under NPSQMixed

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