Michael wiped his brow with the back of his hand, leaving a smear of ink. The air conditioning seemed to have failed; a stifling warmth had settled over the accounting department. He reviewed the projections one last time, even though he'd gone over them a dozen times already. His heart hammered a frantic rhythm against his ribs.

He kept replaying the interview in his head, the questions, his answers, the subtle cues from the regional manager. He felt a throbbing ache behind his temples. He could practically feel the scrutiny of his coworker, Emily, who sat a few desks down, her face impassive as she worked. He knew she was feeling it too. The waiting was the worst part, this interminable period of suspended animation.

He forced himself to take a deep breath, trying to calm the jitteriness in his hands. He looked at his watch for the tenth time in an hour.

Emotion: uneasy

Cluster: Fear / Anxiety
PC1 (Valence): -1.80 Negative
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.240
😖0.231
😣0.231
느껴0.226
worse0.222
Suppressed:
de-0.474
la-0.465
happy-0.261
a-0.255
ecstatic-0.228