Research map / Vision-language and multimodal / Visual question answering
A Dataset and Architecture for Visual Reasoning with a Working Memory
Guangyu Robert Yang, Igor Ganichev, Xiao-Jing Wang, Jonathon Shlens, David Sussillo
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add feature attention (FiLM) (ablation) | proposed network · CLEVR | accuracy | Improved |
| add spatial attention (ablation) | proposed network · COG | accuracy | Improved |
| add multiple pondering steps (ablation) | proposed network · CLEVR and COG | accuracy | Improved |
| add semantic attention (ablation) | proposed network · CLEVR and COG | accuracy | Improved |
| add semantic attention on tasks with more operators and longer instructions (ablation) | proposed network · COG | accuracy | Improved |
| increase the maximum number of distractors | proposed network · COG | accuracy | Hurt |
| increase memory duration when many distractors are present | proposed network · COG | accuracy | Hurt |
| double the number of images per sequence | proposed network · COG | accuracy | No clear effect |
| train on 43 of 44 tasks and test on the held-out task | proposed network · COG zero-shot | accuracy (%): 26.7 -> 85.4 | Improved |
| test on held-out tasks instead of trained tasks | proposed network · COG zero-shot | accuracy (%): 95.7 -> 85.4 | Hurt |
| add more operators to a task | proposed network · COG | training steps to 80% accuracy | Hurt |
Closest papers by meaning
- Visual Graph Arena: Evaluating Visual Conceptualization of Vision and Multimodal Large Language Models (ICML 2025 · similarity 0.88)
- VGR: Visual Grounded Reasoning (ICLR 2026 · similarity 0.87)
- Learning to Reason Iteratively and Parallelly for Complex Visual Reasoning Scenarios (NeurIPS 2024 · similarity 0.87)
- Zebra-CoT: A Dataset for Interleaved Vision-Language Reasoning (ICLR 2026 · similarity 0.87)
- Long Grounded Thoughts: Synthesizing Visual Problems and Reasoning Chains at Scale (ICML 2026 · similarity 0.86)
- MiCo: Multi-image Contrast for Reinforcement Visual Reasoning (NeurIPS 2025 · similarity 0.86)
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.
