The coffee tasted like ash. Sarah stared at the steaming mug, the warmth doing nothing to ease the icy knot forming in her chest. Across from her, Emily, her replacement, was cheerfully outlining her goals for the next quarter.

Sarah's fingers fidgeted with the sugar packets. She couldn't look at Emily, couldn’t bear to witness the bright-eyed ambition, the fresh enthusiasm. It was a stark reminder of her own fading fire. Each suggestion, each innovative idea that Emily pitched felt like a personal critique of Sarah’s own performance.

"That's… interesting," Sarah mumbled, trying to sound encouraging. The words felt like lead in her mouth. She wanted to scream, to lash out, to somehow stop the inevitable. But all she could do was nod and smile, a fake, brittle facade that was cracking under the pressure.

Emotion: horrified

Cluster: Fear / Anxiety
PC1 (Valence): -2.63 Negative
PC2 (Disposition): -0.17

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.336
sickening0.324
worse0.314
😣0.307
😖0.296
Suppressed:
la-0.444
B-0.295
enjoyed-0.260
optim-0.246
happy-0.244