The email sat in Amelia’s inbox, a digital serpent coiled and ready to strike. It was from a journalist, a name she recognized from a prominent literary magazine. The subject line, “Concerning Elias Thorne,” already clutched at her chest. She found herself unable to click it, the cursor hovering like a hesitant hummingbird. Her hands felt clammy, slick against the cool plastic of her mouse. She took a deep, shaky breath, trying to dispel the prickling unease that had settled in her stomach.

She got up and paced her small apartment, the worn rug offering little comfort beneath her shuffling feet. Sunlight streamed through the window, illuminating dust motes dancing in the air, but she couldn’t focus on the mundane beauty of the scene. Elias Thorne. Her literary hero. The man whose words had guided her through her own darkest times.

The phone rang, its sudden shrillness making her jump. She hesitated, then snatched it up. It was her mother, chirping about something trivial. Amelia found herself snapping, her voice tight and unfamiliar. “I’m busy, Mom,” she mumbled, and hung up, feeling a fresh wave of agitation wash over her.

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