He slammed the phone down, the sound reverberating through the sterile, hospital room. The nurse, a young woman with kind eyes, jumped. “Everything alright, Mr. Davies?” she asked tentatively. “Perfect,” Leo snapped, his voice tight. “Just… fantastic.”

His palms were slick with sweat, his jaw locked. He felt a pressure building in his chest, a heavy, suffocating weight. The diagnosis – the same bizarre condition that robbed him of his physical equilibrium – it was a burden, a curse. He wanted to shatter something, anything. He glared at the bland artwork on the wall, a pastoral scene that now seemed mocking in its serenity.

He wanted to tear the IV drip from his arm, rip the wires from his chest. He longed to be free. He thought of his sister, her constant cheer, her unwavering belief in his eventual recovery. He hated her at that moment, hated her optimism, hated the thought of her.

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