John felt his jaw clench. The printer was jammed *again*. He'd only just managed to unstick it, and now it was mocking him with an error message. He kicked the base, the metal groaning in protest.

"Having some trouble, John?" Emily asked, strolling up. Her cheerful tone grated on his nerves.

He glared at her. "No, Emily. I'm just enjoying a leisurely afternoon with this temperamental piece of junk."

Her expression shifted, and she tilted her head. "Wait, were you at Camp Evergreen when you were a kid?"

He almost choked on his own spit. It was like a forgotten memory, a poorly remembered song, tried to bubble up. "Why do you ask?" he snapped, his voice rough.

"I think I saw you there. Didn't you get sent home from the talent show because you kept throwing mud pies at the judges?" Emily seemed to be enjoying herself at his expense. He needed to get out of here.

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