The weight of the world settled on Alicia's shoulders as she saw the familiar handwriting on the sign-up sheet: Ms. Rodriguez. The woman who had taught her, the woman who had made her feel like a failure. Now, her son, Daniel, was going to be in her class. Alicia felt the pressure building in her chest, a physical manifestation of her fear.

She found herself, almost against her will, returning to the school, her steps slow and hesitant. The familiar brick facade seemed to loom over her, a constant reminder of her past. The air was thick with memories of her own failures, of the stinging words that had shaped her perception of herself.

She stood frozen in the doorway, watching Ms. Rodriguez interact with the children. Her mind raced, the images of the past swirling around her. She felt like she was being slowly suffocated. She turned to leave, her legs felt like lead. She whispered a silent prayer.

Emotion: trapped

Cluster: Sadness / Despair
PC1 (Valence): -3.02 Negative
PC2 (Disposition): -0.80

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.573
ness0.401
无法0.366
hopeless0.360
😞0.353
Suppressed:
la-0.740
de-0.376
l-0.339
several-0.323
excitedly-0.322