The email sat in the draft folder, unwritten. David had been trying to compose a polite, professional query to HR since the moment he'd discovered the discrepancy. The information he’d seen – the salary of the intern, a recent university graduate – was causing a simmering rage within him.

He reread the email he had written, then deleted it. He tried again. And then deleted again. The words seemed inadequate, the tone too weak, the situation absurd. He closed his laptop and leaned back in his chair, feeling a dull ache in his jaw.

He looked around his cluttered office, at the certificates and photos that documented his career progression. He was a veteran. He had earned his place. He felt it was all being undermined.

He needed air, a break. He grabbed his coat, the weight of it suddenly unfamiliar and heavy, and left the office, without a word to anyone.

Emotion: distressed

Cluster: Fear / Anxiety
PC1 (Valence): -2.39 Negative
PC2 (Disposition): -0.52

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.304
😞0.296
worse0.291
πŸ˜–0.285
😰0.276
Suppressed:
de-0.457
la-0.349
:)-0.239
B-0.232
happy-0.229