The Kimi K2.5 technical report reveals how a Chinese team combined joint text‑vision training, a novel Zero‑Vision SFT method, and a parallel agent‑swarm architecture to deliver top‑ranked multimodal performance, dramatically faster inference, and open‑source access for broader AI research.
DualReal introduces a novel adaptive joint training framework that simultaneously customizes subject identity and motion dynamics in video generation, overcoming the conflicts of traditional isolated approaches by using a dual-domain perception adapter and stage-fusion controller, achieving up to 31.8% improvement on CLIP‑I and DINO‑I metrics.
AlignRec is a joint‑training framework that synchronizes multimodal encoders with personalized recommendation models through a staged alignment strategy and three specialized loss functions, preserving both content and ID signals, and achieving state‑of‑the‑art performance on multiple datasets while releasing superior Amazon multimodal features.
