The notification blinked: "Brad Peterson sent you a friend request." Amelia stared at the screen, a heat rising in her neck. Her fingers twitched, almost involuntarily reaching for the "ignore" button. She pulled them back. Brad Peterson. The bane of her existence in sixth grade, the boy who'd called her "four-eyes" and stolen her lunch money. Now, a friendly little request for connection.

Her breath hitched. She scrolled through his profile, the photos of a grinning, sun-kissed man, a far cry from the awkward pre-teen who'd tormented her. He looked… successful. She felt a prickling sensation in her palms, a sudden dryness.

She slammed her laptop shut. The urge to flee, to disappear, was almost overwhelming. She went to the kitchen and made a cup of tea, the clatter of the teacup against the saucer abnormally loud in the sudden quiet of the room. She kept replaying the memory of him, his smirk and the way he'd always know the right time to make her life miserable.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242