Search papers, labs, and topics across Lattice.
UC Santa Barbara
2
0
3
Retaining visual figures in skill artifacts boosts CUA performance by over 23 points, proving that seeing is believing in agent training.
Forget handcrafted metrics: RetouchIQ uses an RL-tuned MLLM to generate its own reward signals for instruction-based image editing, leading to more semantically consistent and perceptually pleasing results.
