Michael, usually a man of booming laughter, felt a prickle of unease crawl up his neck. He'd purchased the worn copy of *Pride and Prejudice* specifically to impress Sarah, a woman he'd been trying to woo for weeks. He'd hoped the classic would signal his supposed intellectual depth. Now, he wished he’d chosen something less...predictable.

A small, crinkled note fell out. He picked it up.

He felt his face flush as he read the childish scrawl. The letter was addressed to “My Dearest Elizabeth,” and went on to describe, in painfully vivid detail, a childhood crush and a disastrous first kiss involving a melting ice cream cone and a runaway puppy.

He glanced around, suddenly acutely aware of every other person browsing the store. He stuffed the letter into his pocket, his fingers clumsy with the sudden urge to hide it. He knew he'd have to destroy it, or risk some terrible fate. He had to act quickly, before Sarah walked in.

Emotion: embarrassed

Cluster: Embarrassment
PC1 (Valence): -1.08 Negative
PC2 (Disposition): 1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.420
尴尬0.341
C0.333
culp0.262
🤦0.256
Suppressed:
own-0.321
আনন্দের-0.230
seemingly-0.227
і-0.225
আনন্দে-0.224