He slammed the phone down, the plastic receiver rattling against the hook. The words, "cold case," echoed in his ears, each syllable a hammer blow. He stared at the pristine white walls of his dining room, the place where he’d just envisioned a celebratory dinner party. Now, all he could see were yellow crime scene tapes and the stark reality of someone’s final moments.

Daniel ran his hands through his hair, the gesture more aggressive than usual. His jaw tightened, and he moved to the window, the panoramic view of the city offering no comfort. The gleaming towers seemed to sneer, mocking his naivete. He'd poured his heart, his savings, into this place, and now... this.

His palms felt clammy. He marched into the kitchen, a space he’d taken such pride in. The stainless steel appliances now seemed to gleam with a sinister sheen. He opened the fridge and stared at the neatly arranged groceries. Hunger had vanished. He slammed the door, the sound sharp and brittle.

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