AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Disentangling Pictorial Cue Understanding from Language Bias in VLMs via Depth Ordering Task

Yiqian Liu, Iuliia Kotseruba, John K. Tsotsos

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 1 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 few-shot ICL and CoT prompting (ablation)GPT4.1-mini · O3-Daccuracy: 0.5371 -> 0.6045Improved
add few-shot ICL and CoT prompting (ablation)Gemini2.5 · O3-Daccuracy: 0.4728 -> 0.5124Improved
add few-shot ICL and CoT prompting (ablation)DeepSeek-VL · O3-Daccuracy: 0.5264 -> 0.5040Hurt
add few-shot ICL and CoT prompting (ablation)Qwen2-VL · O3-Daccuracy: 0.5248 -> 0.5128Hurt
add few-shot ICL and CoT prompting (ablation)Phi3.5 · O3-Daccuracy: 0.5136 -> 0.5072No clear effect
add few-shot ICL and CoT prompting (ablation)VLMs · Relative Size cueaccuracy: 0.5522 -> 0.6214Improved
add few-shot ICL and CoT prompting (ablation)VLMs · Familiar Size cueaccuracy: 0.5379 -> 0.5021Hurt
add linear perspective cue to height-in-plane and light-shadow (ablation)VLM mean · O3-D HP/LSaccuracy: 0.5444 -> 0.6100Improved
add linear perspective cue to height-in-plane and focusness (ablation)VLM mean · O3-D HP/FOaccuracy: 0.5405 -> 0.5193Hurt
add more depth cues (0-cue to 2-cue images)VLM mean · O3-Daccuracy: 0.4865 -> 0.5282Improved
add more depth cues (0-cue to 2-cue images)DepthAnythingV2 · O3-Daccuracy: 0.5177 -> 0.7446Improved
increase referring clarity from low to highestVLM mean · O3-D syntheticaccuracy: 0.4807 -> 0.5068Improved

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