Search papers, labs, and topics across Lattice.
7
0
12
8
This technical retrospective offers a different account of the CosyVoice lineage, from CosyVoice through CosyVoice 2 and CosyVoice 3 to Qwen-Audio-3.0-TTS, and formalizes this history through four interface dimensions---representation, ownership, availability, and gradient reach---and separate within-paper evidence from cross-paper comparison.
Optimizing attacker sequences can dramatically enhance the effectiveness of hard-label black-box text attacks, outperforming traditional methods by leveraging a strategic combination of attack strategies.
The J-lens reveals that language models operate with a surprisingly sparse causal structure, concentrating energy in specific pathways to predict future outputs.
StreamFlow achieves a remarkable 67.73% accuracy in streaming video understanding while cutting latency and memory usage by over 50%.
Hybridizing autoregressive and speculative decoding, BALANCE boosts task throughput in edge LLM inference while managing latency and memory constraints.
GROW achieves a 22.7% reduction in word error rate while accelerating training by 2.9x, redefining efficiency in TTS reinforcement learning.
