Search papers, labs, and topics across Lattice.
3
0
4
0
SLM's low-dimensional emotion-specific subspace outperforms CFM's entangled representations, revealing critical trade-offs in emotion steering for TTS models.
Current LALMs exhibit significant performance disparities across cognitive auditory capabilities, revealing a critical oversight in existing evaluation methods.
LALMs struggle to remember non-speech sounds across multi-turn conversations not because of faulty attention, but because their internal representations of those sounds drift over time.
