The weight of the world felt particularly heavy today. He had endured a series of setbacks, and the emotional burden was almost unbearable. He had finally made the call, booking an appointment with a therapist, hoping for some semblance of relief. The sterile environment of the office only amplified his sense of isolation.

A woman opened the door and offered him a warm smile. "Mr. Peterson?" it was Angela, who had tormented him throughout middle school. Her eyes held an understanding that only seemed to intensify his inner turmoil. He felt a sudden, sharp pain behind his eyes.

"I need to go," he blurted out, turning to flee.

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