The clatter of silverware against the fine china felt like a personal attack. Aunt Carol's announcement – that she'd embraced Buddhism – hung in the air, thick and cloying like burnt caramel. Sunlight streamed through the dining room window, illuminating dust motes that seemed to dance mockingly. He gripped his water glass so tightly, he was surprised it didn't shatter. The forced smiles and polite congratulations from the rest of the family grated on him. He wanted to scream, to overturn the table, to make them *feel* something other than this placid acceptance.

He pushed his chair back, the legs scraping against the polished floor, a sound far too loud. Excuses. He needed excuses. "The, uh, parking meter," he mumbled, his voice a low growl. "Better go check it." He escaped the suffocating cheerfulness, the saccharine smiles, and fled.

Outside, the late afternoon sun beat down, and he welcomed the sting of the heat on his skin. He kicked a discarded soda can across the pavement, the metallic clang echoing in the sudden silence. His jaw clenched. Why? Why would she do this?

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