Search papers, labs, and topics across Lattice.
4
0
11
1
Sorting prompts by their reliability can dramatically enhance the effectiveness of on-policy distillation, leading to superior performance in complex tasks.
Recursive self-improvement in Macaron-V1 leads to continual learning that adapts to real-world experiences, setting a new standard for open agent models.
Shifting the focus from photorealistic rendering to dynamic visual changes, DC-WAM enhances robot policy performance by 20% in challenging environments.
Finally, a voice design model that can handle both single utterances and multi-turn dialogues with improved expression controllability and contextual awareness.
