Search papers, labs, and topics across Lattice.
3
0
5
Training-free image editing can now beat fine-tuned models, thanks to a clever way of injecting visual context and ensuring concept alignment.
Ditch the multi-camera setup: PrismMirror synthesizes photorealistic frontal views of humans from a single image in real-time, outperforming prior methods in both visual quality and geometric accuracy.
Fine-tuning LLMs doesn't have to compromise safety: TOSS identifies and removes unsafe tokens with greater precision than sample-level methods, preserving utility.
