Search papers, labs, and topics across Lattice.
4
0
7
Subjective rewards in speech generation are not interchangeable, revealing critical nuances in how RL aligns with human listener preferences.
Unigram entropy emerges as a key predictor of codec performance, revealing stark differences in how various architectures handle noise conditions.
APT accelerates high-resolution Diffusion Transformers by leveraging attention probabilities, achieving unprecedented speedups and energy efficiency in generative AI tasks.
Korean models struggle significantly with writing-system-intensive tasks, revealing a 68.7 pp accuracy gap in the Korean Cipher compared to English.
