The manuscript sat on the bedside table, a constant reminder of the life she’d been living. The main character in her partner’s novel was "Clara," and the parallels were uncanny. Everything from her favorite scent to the way she chewed her gum was in there, immortalized. It was as though he had taken her apart, piece by piece, and then reassembled her, this time in print.

Her usual morning routine, once a source of comfort, now felt like a performance. She went through the motions, brushing her teeth, applying makeup, choosing an outfit, but it all felt hollow. The house seemed to shrink around her, each room closing in a little more.

She tried to read, but the words swam before her eyes. The novel's plot, mirroring her own life, felt inescapable, a labyrinth with no exits. She felt a growing sense of unrest.

"Honey, you haven't touched your coffee," Miles said, his voice laced with concern. She took a sip, the familiar bitterness doing nothing to soothe the tension knotting her stomach.

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