The email sat in Eleanor’s inbox, a shimmering beacon of false hope. Her article, the one she’d bled over for months, was live. Except, it wasn't *hers*. It bore the name of her supervisor, Mark, who’d been vaguely supportive, but whose writing style was about as nuanced as a brick. She reread the article, the words mocking her with their blatant, unearned praise.

A coldness settled in her stomach, a leaden weight. She closed her laptop with a snap, the sound echoing in her otherwise silent apartment. The city outside her window, usually a symphony of sounds, seemed distant, muffled. Eleanor walked to the kitchen, poured a glass of water, and found she couldn't bring herself to drink it. The glass sat untouched on the counter.

Later, she found herself pacing. She needed air, but the thought of leaving the apartment, facing the world, choked her. The space around her felt increasingly small, the walls closing in. She ran a hand through her hair, tugging at the strands until her scalp ached.

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