The email arrived mid-lecture, a stark white rectangle in the midst of Professor Davies’ passionate discussion of Romantic poetry. Sarah’s breath hitched. She surreptitiously opened it under the cover of her textbook, the words – “Urgent Announcement: College Closure” – immediately seizing her attention. Her stomach performed a clumsy somersault, and she felt a sudden chill, despite the stuffy room. She had to clench her jaw to stop the tremor that threatened to break through.

She stared, unseeing, at the professor, whose words now seemed to bounce off a soundproof wall. Her gaze flickered to the faces of her classmates, oblivious, scribbling notes, and a pang of isolation hit her. She felt a need to flee, to be somewhere, anywhere, other than where she was.

The lecture ended, but Sarah remained rooted to her seat, a statue of dread. She slowly gathered her things, her fingers clumsy and hesitant. The weight of her backpack suddenly felt unbearable. Leaving the classroom, she drifted down the hall, avoiding eye contact with anyone, a silent plea for reprieve etched onto her face.

Emotion: worried

Cluster: Fear / Anxiety
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.285
S0.279
😰0.277
😣0.269
Worse0.250
Suppressed:
la-0.476
de-0.432
🎉-0.233
happy-0.231
"-0.228