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.
NGS-Marker effectively thwarts partial infringement in 3D Gaussian Splatting, ensuring robust copyright protection for 3D assets.
SquareCB.Comb achieves minimax optimal regret while scaling efficiently to large combinatorial action sets, redefining the landscape of contextual combinatorial semi-bandits.
