Search papers, labs, and topics across Lattice.
3
3
5
The DeepSeek-V4.1-Flash model, a multimodal Mixture-of-Experts model with 552B backbone parameters and support for contexts of up to one million tokens, is introduced, substantially improving cost efficiency for agentic workloads and pushing the limits of KV cache compression.
GlyphAnchor achieves unprecedented text fidelity in image generation, even with complex and rare characters, by anchoring glyphs to their positions in the target image.
Instruction-based image editing just got a whole lot better: FireRed-Image-Edit leapfrogs existing systems with a massive, meticulously curated dataset and a suite of training innovations.
