The elevator ride felt interminable. As the doors opened, Carol stepped out, her legs shaky. She felt a cold dread settle in her stomach. She had just spoken with HR, and the numbers they presented were, frankly, insulting. New hire, Jake, with less than six months experience, was being paid more.

She walked slowly towards her desk, her shoulders hunched. The open-plan office, usually a space she found energizing, now felt like a vast, empty wasteland. She avoided eye contact with her colleagues, not wanting to betray her inner turmoil.

Her hands fumbled with her keyboard, the familiar clack of the keys grating on her nerves. She felt a tightness in her chest, a physical manifestation of the anger she was trying to suppress. Her mind raced, questioning her value, her contributions, her very worth within the company.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236