The book lay open on the coffee table, a well-worn copy of “The Little Prince” in French. Sarah walked into the living room, expecting to find David watching football. Instead, he was hunched over the book, moving his lips as he read aloud in a halting, accented French. Her cheeks burned.

She’d spent years perfecting her French, a language she cherished, that represented a significant part of her identity. And he… he was stumbling through it, butchering the pronunciation of words she loved. It was… embarrassing, for reasons she couldn’t quite articulate.

David looked up, startled. “Oh! Hey! I was just… trying to… read.” He gestured awkwardly at the book. She averted her gaze, suddenly noticing a rogue thread on the sofa.

“Right,” she replied, her voice flat. She moved to the kitchen to make some tea, the silence filling the space between them.

Emotion: mortified

Cluster: Embarrassment
PC1 (Valence): -1.36 Negative
PC2 (Disposition): 0.79

Role in Research

This story is one of 1,000 stories generated for the emotion mortified. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 mortified stories, after denoising with neutral dialogue baselines, produces the mortified emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the mortified vector is projected through the unembedding matrix.

Promoted:
l0.406
尴尬0.257
C0.252
worse0.245
0.242
Suppressed:
own-0.340
了他-0.218
他对-0.203
-0.194
楽し-0.193