The weekly support group meeting was the last place Amelia wanted to be, but her therapist insisted. She sat in the back, shrouded in the shadows, hoping to remain invisible. The low hum of voices and the scent of stale coffee triggered a wave of nausea.

As the meeting progressed, people shared their struggles, their triumphs, their fears. Amelia shrunk further into herself, her hands clasped tightly in her lap. How could she possibly open up, confess her vulnerability? The very thought made her skin crawl.

When a woman, a stranger, started talking about her challenges, her words resonating in a manner that mirrored Amelia's own experiences, Amelia’s breath hitched. But the other woman’s face reflected a similar unease. Their eyes met, and a silent, unwanted understanding passed between them.

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