Search papers, labs, and topics across Lattice.
Carizon, Beijing 100094, China
4
0
7
0
Achieve near-lossless performance in autonomous driving VLMs with 90% token reduction – without any training.
Diffusion models can beat discriminative classifiers at facial expression recognition, but only with a dynamically adjusted margin loss that accounts for per-sample difficulty.
Forget generic fine-tuning data — Bloom's Taxonomy-based data generation can boost LLM performance in complex engineering domains like space situational awareness by up to 176%.
Mamba's linear complexity lets you ditch computationally intensive graph attention for flexible job shop scheduling, achieving state-of-the-art performance with faster solving speeds.
