Research map / Vision-language and multimodal / Human and embodied modeling
Thinking in PolAR Pictures: Using Rotation-Friendly Mental Images to Solve Leiter-R Form Completion
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| use three subslices instead of two (ablation) | PolAR-SS · Leiter-R Problem 09 | slice localization accuracy | Helped |
| two subslices only (line segment) on rectangle-dominated problem (ablation) | PolAR-SS2 · Leiter-R Problem 09 | slice localization accuracy | Hurt |
| pixel-based Max-Min similarity under scanning color differences (ablation) | all PolAR models · Leiter-R Problems 04, 05, 08 | slice localization accuracy | Hurt |
| transparency analysis on slices with unenclosed white pixels (ablation) | PolAR-S and PolAR-H · Leiter-R Problem 14 | slice localization accuracy | Hurt |
| use subslice-based search that avoids transparent pixels (ablation) | PolAR-SS · Leiter-R Problem 14 | slice localization accuracy | Helped |
| lenient choice-level metric (at least one slice located) instead of slice-level metric | all PolAR models · Leiter-R Form Completion | localization accuracy | Helped |
| noisy but unscaled, unshaded slices (copies of slot subregions) | all PolAR models · Leiter-R hard sample multi-slot problem | localization accuracy: 100% | Helped |
| scaling and shading differences between slices and easels (ablation) | all PolAR models · Leiter-R Problems 12/13 | choice localization accuracy | Hurt |
Closest papers by meaning
- Spatial-IQ: Deconstructing Spatial Intelligence via Hierarchical Capability Tests (NeurIPS 2026 · similarity 0.85)
- KiVA: Kid-inspired Visual Analogies for Testing Large Multimodal Models (ICLR 2025 · similarity 0.85)
- Hyperphantasia: A Benchmark for Evaluating the Mental Visualization Capabilities of Multimodal LLMs (NeurIPS 2025 · similarity 0.85)
- HumanPCR: Probing MLLM Capabilities in Diverse Human-Centric Scenes (ICLR 2026 · similarity 0.84)
- When Visualizing is the First Step to Reasoning: MIRA, a Benchmark for Visual Chain-of-Thought (CVPR 2026 · similarity 0.84)
- Cross-modal Associations in Vision and Language Models: Revisiting the Bouba-Kiki Effect (NeurIPS 2025 · similarity 0.84)
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.