“And you were such a good boy,” his grandmother had always said. He'd lived his life believing it. Thomas stared at the photographs, the image of his grandfather, a man he'd only known through stories, grinning with a group of shady looking men.

He felt a tightness in his chest, a sensation akin to being physically squeezed. He ran a hand through his hair, the frustration simmering inside him.

He slammed his fist on the table, the force rattling the dishes. He wanted to shout, to rail against the deception. The comfortable world he thought he knew was suddenly fractured and unrecognizable.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290