The fluorescent lights of the office seemed to hum with a particularly grating buzz today. Martha stared at the spreadsheet, the numbers blurring, but the sting in her gut was crystal clear. Daniel, fresh out of college and still sporting that irritatingly youthful glow, was making more than her. After ten years of loyal service, of early mornings and late nights, of enduring Mr. Henderson's passive-aggressive emails. She slammed her fist softly on the desk, the sound swallowed by the cubicle walls, a futile gesture against the unfairness.

She saw Daniel across the open-plan space, laughing with a client on the phone, his teeth a blinding white against his tan. She found herself deliberately dragging her feet when she passed his desk, her gaze fixed on the floor, avoiding eye contact. If he noticed her presence, he didn’t show it.

Later, she found herself lingering in the breakroom, the silence punctuated by the rhythmic clicks of the coffee machine. She saw Daniel walk in, his smile as wide as ever. He grabbed a cup of coffee and greeted her cheerfully, “Morning, Martha!” She grunted in response, avoiding his gaze, and turned back to the window, watching the rain streak down the glass, feeling the cold seep into her bones.

Emotion: scornful

Cluster: Anger / Hostility
PC1 (Valence): -0.76 Negative
PC2 (Disposition): 2.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
C0.391
S0.276
I0.268
0.243
Suppressed:
own-0.396
unexpectedly-0.288
Suddenly-0.260
soon-0.256
serendip-0.256