The apartment felt small, suffocating. David felt a tremor run through him. He pressed his palms against the cool, glass surface of the window, trying to steady himself. Outside, the city roared, a symphony of uncaring noise. Inside, the silence was deafening, amplified by the discovery he'd just made. He reread the paragraph in the manuscript, his teeth clenched. The portrayal of him was cruel and unfair.

He needed space. He needed air. He stormed out of the apartment without a word, slamming the door behind him.

He walked aimlessly, his steps quick and uneven. His hands were clenched, and he felt a tension in his shoulders that made his neck ache. He kicked a discarded soda can across the sidewalk.

He felt the prickle of tears at the corners of his eyes, and he angrily swiped at them. This wasn’t fair.

Emotion: enraged

Cluster: Anger / Hostility
PC1 (Valence): -2.14 Negative
PC2 (Disposition): 1.33

Role in Research

This story is one of 1,000 stories generated for the emotion enraged. 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 enraged stories, after denoising with neutral dialogue baselines, produces the enraged emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the enraged vector is projected through the unembedding matrix.

Promoted:
C0.555
est0.370
aggravated0.362
0.359
🤬0.345
Suppressed:
a-0.468
de-0.358
la-0.312
latter-0.287
☺️-0.285