The sterile white walls of the hospital waiting room seemed to press in on Emily. She fidgeted, adjusting her handbag, tracing the pattern of the floral wallpaper with her finger. Across from her, Mark, her friend since university, sat with a composure she couldn’t muster. The interview was over, but the results were still pending.

She had envisioned this job for so long, a chance to get out of her cramped flat. She had dreamed of a fulfilling career, of making a real difference. The silence in the room felt suffocating, punctuated only by the rhythmic beeping of medical equipment from the adjacent rooms. Her heart hammered against her ribs.

She was unable to sit still.

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