The smell of burnt popcorn still clung to the air. Sarah had attempted a movie night, a small gathering of friends to celebrate a new job. Turns out, the movie was just a backdrop to the real event: the heated argument between her and Mark, her boyfriend. Now, the super was at her door.

He looked around, taking in the scene: the half-empty bottles of wine, the overturned cushions, the lingering scent of smoke. He’d never been the friendliest man, but his usual gruff demeanor was replaced by a look that was almost pitying.

Sarah’s face flushed. Every square inch of her skin felt like it was on fire. She felt a burning need to explain, to apologize, to somehow make amends. But the words wouldn't come. She just stood there, mortified, as he shook his head and walked away, muttering something about noise complaints.

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