The world blurred around him. David stared, disbelieving, as the tow truck started to pull away. His blood felt like ice water in his veins. He ran after it, shouting incoherently, but his voice was swallowed by the morning air.

He’d paid his taxes, kept his nose clean. Now, this. He felt a searing heat build in his chest, radiating outwards. When he finally stopped, gasping for breath, he kicked the curb so hard his toes throbbed. He turned and stomped back to the house, his fury a tangible presence.

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