Moshi全双工语音模型情绪激活引导研究
Get the story
该研究在全开源全双工语音语言模型Moshi中探索情绪转向,采用均值差激活转向覆盖四种情绪,每帧只需几次向量加法、无需重训。研究发现情绪可从残差流线性解码,但激活转向仅部分可行且不均衡:happy、angry与surprise共享同一方向,sad则可独立转向,且三者共享成分无法从所有情绪中等量投影剔除。论文提出情绪转向遵循几何结构而非标签,并被NeurIPS 2026实时会话智能体研讨会(RTCA)接收。
Generated from reports · updated 3 hr ago
Timeline
Follow the coverage from different angles.
- arXiv · Computation and Language情绪转向遵循几何结构而非标签:全双工语音模型中的情绪方向
研究在全开源全双工语音语言模型 Moshi 中探索情绪转向,用均值差激活转向覆盖四种情绪,每帧只需几次向量加法、无需重训。情绪可从残差流线性解码,但激活转向仅部分可行且不均衡:happy、angry 与 surprise 共享同一方向,sad 则可独立转向,且三者共享成分无法从所有情绪中等量投影剔除。论文被 NeurIPS 2026 实时会话智能体研讨会(RTCA)接收。
Heat trend
Current heat 9·Comparable peak 10(Oct 8)·Comparable change over 24 hours –
The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.