The library felt like a mausoleum. Sunlight streamed through the tall windows, illuminating dust motes dancing in the air, but the effect was more oppressive than uplifting. Amelia traced the familiar lines of her college essay, now a highlighted example in the admissions guide. A giggle bubbled in her throat, quickly stifled. She closed her eyes, imagining the thousands of eyes that had read this, the words that revealed the deepest parts of her. She felt exposed, a specimen pinned to a board. The weight of expectations settled heavily in her chest, a physical ache.

Her fingers drummed a frantic rhythm against the glossy page. This essay, her voice, had been a key to escape, the only thing that felt like freedom. Now, it was a chain. She had to apply to graduate school, and this essay followed her like a phantom limb. The thought made her skin crawl.

She considered tearing the book. Then what? Return to the suffocating familiarity of her hometown, and her parent's relentless pressure to settle down? The thought of that routine made her stomach churn.

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