“So, you bought the Murder House? Wow,” her friend, Janice, said with a low whistle. The casual comment, made over coffee, sent a jolt of heat through Elena. She wished Janice would just shut up. The news reports had mentioned it, of course, the gruesome details… The kitchen, where she made her morning coffee, the very place of the crime.

Elena felt the urge to shift, to fidget. She took a long, slow sip of her lukewarm coffee, trying to appear nonchalant. But Janice’s knowing look, the way she was clearly savoring the story, made her skin prickle. She forced a laugh, a brittle, unnatural sound. This was going to make hosting her book club a lot harder than she thought.

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