Liam stared at his phone screen, a knot tightening in his chest. His thumb hovered over the ‘block’ button, but he hesitated. Mr. Davies, his art teacher, followed him on Twitter. Liam felt a coldness wash over him. He had tweeted a string of sarcastic comments about the school's new art curriculum, not realizing his teacher might see them.

He slumped onto his bed, burying his face in his hands. The air in his room felt suddenly thick, suffocating. He’d poured his frustration into those tweets, thinking he was safe behind his screen. Now, he felt like a child caught with his hand in the cookie jar. He didn’t want his teacher knowing how bored and uninspired he was in the class.

A dull ache throbbed in his temples, and he avoided eye contact with the poster of Monet on his wall.

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