The email, subject line "Important Building Update," hung on her screen. Sarah read it, re-read it, then slammed her laptop shut with a decisive click. The small act didn’t satisfy the surge of frustration that coursed through her. She felt a buzzing in her ears, a pressure behind her eyes that made her blink rapidly.

She grabbed a half-eaten bag of chips and ripped it open, then began to eat them with an almost violent energy. Salt crumbs scattered across her desk. She’d always loved this apartment, its sunny windows and quirky layout. She hated the thought of losing it.

She glared at the empty, lifeless screen, picturing all the things that would disappear: the friendly neighbors, the worn wooden floors, the sense of belonging. The apartment had been a constant in her life, a safe harbor. Now, she felt a profound sense of instability.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210