Microsoft released Mage-Flow, a 4-billion-parameter image generation model that aims to match the quality of larger models while being smaller. The model uses a co-design approach with Mage-VAE, a lat…
A developer trained the same two faces on six base models (Ideogram, Flux.1 Dev, Flux.2, Klein, Krea, Z-Image) to find the best base model for character LoRA training in 2026, concluding that Ideogram…
Researchers introduced i2L (image-to-LoRA), a framework that compresses style LoRA training into a single forward pass, enabling immediate style instantiation from reference images without per-style o…
Researchers have developed Diff-Instruct with Diffused Reward (DIDR), a data-free trajectory-level alignment framework that propagates reward-tilted clean-image distributions across all noise levels t…