Unverified50% confidenceFactExact time
After removing the acoustic tokenizer, the model must model directly in a continuous acoustic space, typically relying on continuous generation paradigms such as diffusion models or flow matching, which shortens the inference pipeline.
1
Sources
50%
Confidence
Long-term
Relevance
10/23/2026
First Seen
Sources
Related Entities
Related Claims
Unverified传统大模型生成是自回归的，必须生成完整序列才能输出，而实时语音场景要求模型在接收输入的同时开始生成输出73% similarUnverifiedThe background reasoning mechanism decouples time-consuming reasoning computation from the real-time audio stream, allowing the model's voice response layer and reasoning computation layer to run in parallel73% similarUnverified端到端音频模型本质上是一个自回归生成模型，预测下一个音频token时依据此前所有音频token的概率分布73% similarUnverified流式推理技术让模型对持续到来的音频帧进行增量式处理，在达到足够置信度时立即开始输出，而非等待说话人完整停顿73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1089328API
curl https://kongchang.com/api/v1/knowledge/claims/1089328MCP
get_claim(id=1089328)