She found herself pacing the length of her small office, the click of her heels echoing in the sterile space. Sunlight streamed in, highlighting dust motes dancing in the air, but the scene did nothing to calm the internal churn. A notification blazed on her laptop screen – a Facebook friend request from Jason. The memory of his cruel laughter, of the way he'd tripped her in the cafeteria, of the whispered insults, flooded back. Her palms were clammy. She pressed her lips into a thin line, the corners of her mouth twitching. The air suddenly felt too thick, too heavy.

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