Search papers, labs, and topics across Lattice.
KAIST AI
3
0
8
Get persistent world memory in your video diffusion model for free: WorldKV matches full KV-cache fidelity at 2x throughput without any training.
Architectural specialization and an efficient training approach enable Motif-Video 2B to outperform larger models with a fraction of the resources.
Stop relying on keyword matching: FinTexTS uses semantic similarity and LLM-driven multi-level classification to pair financial text with time-series data, boosting stock price prediction.
