The notification buzzed, a tiny electric shock. *Michael Davies*. He felt a wave of dizziness. He grabbed the desk for support. His vision blurred, and the office around him seemed to shrink, the walls closing in. He felt the familiar weight in his chest.

He forced himself to take a deep breath, fighting the urge to bolt from the room. He felt his heart slamming. He had to stay calm. He had to be professional. This was the era of social media, and he could not afford to be unprofessional.

He took another shallow breath and reached for his water bottle, and chugged the entire contents down in one go. He slowly looked back at the screen, and the friend request stared back at him.

Emotion: panicked

Cluster: Fear / Anxiety
PC1 (Valence): -2.44 Negative
PC2 (Disposition): 0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.317
😫0.294
😖0.285
😰0.285
escape0.283
Suppressed:
la-0.432
a-0.374
de-0.296
楽し-0.262
happy-0.258