"So, how's the new car, Daniel?" The question from his boss, Mr. Henderson, hung in the air, heavy with unintended irony. Daniel had been driving his ten-year-old clunker, while the freshly minted graduate, Mark, was now sporting a sleek sports car. Daniel felt his cheeks flush.

He mumbled something about the car being reliable, his voice barely a whisper. He found himself avoiding Mark in the hallways, strategically timing his coffee breaks to avoid any chance encounters. The image of the shiny new vehicle was a constant, unwelcome reminder of his own perceived inadequacy.

He focused on his work, throwing himself into projects with a renewed, desperate intensity, hoping to somehow compensate. He skipped lunch, unable to face the social dynamics.

Emotion: embarrassed

Cluster: Embarrassment
PC1 (Valence): -1.08 Negative
PC2 (Disposition): 1.10

Role in Research

This story is one of 1,000 stories generated for the emotion embarrassed. 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 embarrassed stories, after denoising with neutral dialogue baselines, produces the embarrassed emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:
l0.420
尴尬0.341
C0.333
culp0.262
🤦0.256
Suppressed:
own-0.321
আনন্দের-0.230
seemingly-0.227
і-0.225
আনন্দে-0.224