"Seriously?" Chloe’s voice was clipped, sharp. The accusations – veiled, yet pointed – from the TA had her fingers tapping a frantic rhythm on the table. She could feel her pulse thrumming in her temples. The air in the small office felt dense, oppressive. It was as if her own brain was trying to claw its way out of her skull.

She wanted to laugh, to dismiss it as absurd, but the look on the TA's face, a mixture of pity and disapproval, prevented it. She felt a heat rise in her chest, a need to lash out, to defend herself. Instead, she took a shaky breath and asked, "Can you please be more specific?" Her voice was a strained whisper.

Emotion: irritated

Cluster: Anger / Hostility
PC1 (Valence): -0.86 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.370
S0.282
P0.264
😠0.248
aggravated0.233
Suppressed:
own-0.300
joyful-0.211
joyous-0.207
खोले-0.206
newfound-0.205