GQA

GQA is a new dataset for real-world visual reasoning and compositional question answering, seeking to address key shortcomings of previous visual question answering (VQA) datasets.

Papers using GQA (8)

K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model2026A Fast, Reliable, and Secure Programming Language for LLM Agents with Code Actions2025ProTo: Program-Guided Transformer for Program-Guided Tasks2021 · 15 citesModular Visual Question Answering via Code Generation2023Recursive Visual Programming2023Pyramid Coder: Hierarchical Code Generator for Compositional Visual Question Answering2024Modular Visual Question Answering via Code Generation2023K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model2026