Her fingers, usually deft and quick, fumbled with the clasp of her purse. A vein pulsed in her forehead. "You did *what*?" Sarah asked, her voice a brittle weapon. The cafe's cheerful chatter felt invasive, the sunlight too bright.

Across the small table, Mark fidgeted, his smile weak. "Just… a few lessons in Gaelic. For fun."

Fun? The word grated. She could barely manage to say two words in her native tongue, and Mark, who struggled with basic English grammar, thought it was ‘fun’ to delve into the complexities of her heritage? She glared.

She took a deep, shaky breath, fighting the urge to stand up and leave. “So you’ve been mocking me, haven’t you?” she heard herself whisper. The coffee suddenly tasted like ash in her mouth.

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.430
🤬0.350
aggravated0.343
0.340
0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253