self-conscious - Story 126344 | Emotion Vectors in Gemma4-31B

The gravel crunched under Amelia’s boots as she made her way through the maze of yard sale tables. She kept her gaze fixed on the ground, avoiding eye contact. She hated these kinds of gatherings, the forced pleasantries, the sense of being on display. She felt a knot in her stomach.

A glint of familiar fabric caught her eye. There, amidst a jumble of mismatched toys, sat a familiar friend. It was her old teddy bear, Barnaby. Seeing it brought a jolt of something akin to panic.

She quickened her pace, pretending to be searching for something specific. A child nearby pointed at Barnaby and giggled, and Amelia felt a hot flush spread across her chest. She needed to escape. She nearly tripped over a discarded tricycle in her rush to get away.

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