The notification popped up: "Mark Thompson sent you a friend request." Sarah’s breath hitched. She nearly dropped her phone, the screen flashing the familiar Facebook blue against her trembling fingers. Mark Thompson. The same Mark Thompson who used to trip her in the school hallway, steal her lunch money, and call her names she still remembered. A heat flushed her cheeks.

She considered ignoring it, but curiosity, that insidious little devil, won out. She clicked on his profile. He looked…older. Slightly softer around the edges, a few gray hairs. He had a family photo. A wife, two kids. A wave of conflicting feelings washed over her.

She closed her eyes, quickly. Opening them, she typed a quick, "Sure." Then, she shut her laptop and went to the kitchen to busy herself with washing the dishes.

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