He had been working late. Exhausted, he'd sunk into the couch. The book, left carelessly on the coffee table, was now open. It was a novel by his wife, Clara, that he had never known about. Her depiction of him was a bitter pill to swallow.

The words stung. He was depicted as a weak, ineffectual man, clinging to the vestiges of his youth. A flicker of something unpleasant ignited inside him. He started reading the parts that were about him, and found his jaw clench, his hands forming fists.

He snapped the book shut. He got up and began to pace, back and forth, the floorboards groaning beneath his weight. Then he went to the fridge and grabbed a beer, and chugged it down, quickly, and walked back and sat on the couch again. He felt a deep, profound sense of disappointment.

Emotion: angry

Cluster: Anger / Hostility
PC1 (Valence): -2.05 Negative
PC2 (Disposition): 1.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.354
est0.354
aggravated0.342
🤬0.337
0.323
Suppressed:
de-0.424
a-0.411
la-0.341
H-0.320
L-0.296