“How can this be?” thought Sarah. The board in front of her showed another flight delay. She was scheduled to give a lecture at a prestigious academic conference that was due to start in an hour.

She slumped in her seat, feeling the weight of the situation. The sweat ran down her back and she fidgeted, rearranging her clothes. Her carefully prepared presentation, her months of research, all felt suddenly worthless.

She imagined the faces of the audience, the judges, the other speakers. Each thought was a blow. She took a deep breath, trying to regain some semblance of composure. She pulled out her phone and started writing an email to the conference organizers.

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