He stared at the peeling paint of the ceiling, his jaw locked. Each tick of the grandfather clock in the reception area felt like a hammer blow against his skull. It had been ten years since the incident, ten years of nightmares and flashbacks. Yet, here he was.

The door opened, and a woman with an unnervingly familiar smile greeted him. His heart began a rapid tattoo against his ribs. It was her, Emily. He recognized the way her eyes crinkled at the corners.

“Come in, David,” she said, her voice smooth, almost melodic. He knew that tone; it was the one she’d used just before she’d unleash her cutting words in the schoolyard.

He followed her into her office, every muscle in his body rigid. The space seemed to shrink as he took a seat, the chair feeling too close to the floor.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198