Hugging Face Papers ↗ · 2026-11-18
AVTR: Adaptive Multimodal Reasoning with Visual Imagination
Media reports Hugging Face Papers Source language: Chinese
AVTR combines generative visual reflection with textual reasoning and learns to route between them per input, improving multimodal task accuracy while reducing inference latency.
Full text
Adaptive Visual-Textual Reasoning combines generative visual reflection with textual reasoning and learns to route between them per input, improving multimodal task accuracy while reducing inference latency.