Voice model Samuel learns human-like mouth positioning without explicit training

deliprao · x · 2026-09-29

Developer vvolhejn observed that the Samuel voice model often positions its virtual mouth the way a human would, despite never being explicitly trained to do so.

This emergent, human-like articulation suggests multimodal voice models implicitly pick up subtle communication cues. The original post links to a demo video.

Original post →

More from Models

Models channel →