He tried to ignore the congratulatory texts, the emails flooding his inbox. But the article, *his* article, was everywhere. On LinkedIn, Twitter, and the university website, under the name of Professor Evelyn Reed. Daniel felt a wave of nausea wash over him.

He avoided eye contact during the staff meeting, focusing instead on the patterns in the carpet. His cheeks burned, his palms were slick with sweat. He felt as if everyone could see through him, could see the fraudulence of the situation.

Later, he walked to the park, the crisp autumn air offering little relief. He kicked at a pile of leaves, a frustrated grunt escaping his lips. He sat on a bench, head in his hands, the weight of the secret crushing him.

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