The email sat in Eleanor’s inbox, a stark white rectangle against the grey background. Professor Davies’s name, the subject line: “Departure from the University.” Her hands felt clammy, and she ran a finger along the seam of her jeans, trying to find some purchase against the sudden, unsettling looseness in her stomach. The coffee she'd made that morning seemed to have vanished.

She reread the email three times, each pass a frantic search for some hidden meaning, a retraction perhaps, a cruel joke. But no. The words remained fixed, unyielding. He was leaving. Two years of research, countless hours in the lab, all hinged on his guidance. Now what? Her heart thumped a staccato rhythm against her ribs.

Eleanor grabbed her phone, scrolling through contacts, fingers trembling. Should she call her parents? Her roommate? The department head? The thought of forming the words, of making it real, choked her. Instead, she stared out the window, at the swaying branches of the oak tree outside, each leaf a dizzying swirl of green. The impending loss felt like a physical weight, pressing down on her chest.

Emotion: anxious

Cluster: Fear / Anxiety
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.315
😖0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256