The invitation sat on Clara’s kitchen counter, a pristine white rectangle taunting her. Her mother's face, smiling and serene, stared back from the glossy photo attached. Retirement. A life of leisure, of travel, of… what exactly? Clara wasn’t sure, but she knew she wasn’t the picture of success her mother envisioned. She avoided the photo, busying herself with scrubbing the sink until her knuckles were raw. This speech... the thought churned in her stomach.

The night of the party, Clara arrived late, hovering near the back of the room. Every glance, every well-wishing hug, felt like a judgment. Her carefully chosen outfit suddenly seemed too flashy, her smile too tight. When her name was called, her throat constricted. She shuffled to the podium, the microphone feeling impossibly heavy, the words sticking in her mouth like glue.

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