self-conscious - Story 126369 | Emotion Vectors in Gemma4-31B

She squinted at the card in the harsh morning light. A picture of a smiling puppy greeted her. Her shoulders slumped. It was too cheerful. Too forced.

Back in her apartment, she felt an overwhelming urge to clean, as if tidying her home would somehow fix everything. She busied herself with dusting, avoiding the mirror, refusing to make eye contact with her own reflection.

She finally sat down, opening the card with a sigh. "Have a great day!" the message read. Her throat tightened. She felt like everyone was watching, judging her every move. She wished she could disappear.

Emotion: self-conscious

Cluster: Embarrassment

PC1 (Valence): -1.03 Negative

PC2 (Disposition): 0.52

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-conscious vector is projected through the unembedding matrix.

Promoted:

`S`	0.307
`C`	0.298
`ness`	0.273
`尴尬`	0.263
`😶`	0.242

Suppressed:

`que`	-0.399
`de`	-0.266
`triumphant`	-0.266
`আনন্দের`	-0.247
`ايا`	-0.238