The phone vibrated against her desk, a rhythmic pulse that felt like a mocking heartbeat. It was a message from Chris: “Meeting got pushed back. Running late. Xx” Emily glared at the screen, her jaw clenched. He was supposed to be at a conference, a “high-powered” event showcasing his “cutting-edge” software.

She had just stumbled across his Facebook page. The one he’d always kept hidden. It revealed a profile of a part-time waiter and aspiring actor. She felt a wave of dizziness wash over her, a disorientation that stole her breath.

She slammed the phone down. All the little things, the inconsistencies she had ignored, the vague excuses, now felt like a calculated deception. She grabbed her coat, not even bothering to grab her purse. She had to get out, away from the sterile environment of her office.

Emotion: outraged

Cluster: Anger / Hostility
PC1 (Valence): -1.96 Negative
PC2 (Disposition): 1.32

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.467
0.334
🤬0.327
aggravated0.323
0.312
Suppressed:
H-0.315
☺️-0.286
optimistic-0.278
adventurous-0.262
मुस्कुरा-0.260