The small cabin was supposed to be a sanctuary, a refuge from the world. Now, with the biography, "Silas Bellweather: Reclusive Poet" detailing his every move since he'd moved there, it felt like a gilded cage. He stared out at the forest, the towering trees seeming to close in around him, creating a wall.

He slammed his fist on the rough wooden table. The sound ripped through the silence, but brought no relief. The peace he had worked so hard to find was shattered. Every rustle of leaves, every distant bird call, felt like a judgment.

He felt the compulsion to leave, to run, but the idea of facing the world outside, knowing his secrets were laid bare, was more terrifying. He was frozen in place.

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