self-conscious - Story 126611 | Emotion Vectors in Gemma4-31B

The microphone seemed enormous in her hand. Sarah adjusted her glasses, the movement a nervous tic she’d developed over the years. The ballroom, usually filled with laughter and dancing, suddenly felt oppressively silent. Each eye on her felt like a spotlight, highlighting every perceived flaw. Her cheeks were burning, and she could feel a bead of sweat tracing a path down her temple.

She glanced at her mother, who was beaming at her from the head table. The smile seemed encouraging, yet also magnified the pressure. Sarah swallowed, trying to remember the opening line she'd practiced a hundred times.

"Um, hello everyone," she began, her voice a reedy whisper that barely carried to the back of the room. She hated her voice.

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