self-conscious - Story 126799 | Emotion Vectors in Gemma4-31B

The phone rang, and it was a reporter from the local paper. He was doing a follow-up story on the…house. Sarah felt a wave of nausea wash over her. It had been months since she moved in, and she’d thought she’d managed to put the past behind her.

She tried to sound calm, but her voice was a little breathy. She answered the reporter’s questions, each one a fresh stab. Her hand trembled as she hung up the phone. She felt trapped, exposed.

She walked through the house, the walls suddenly seeming to close in. She found herself inspecting every window, every lock. She felt a strong urge to get away, to disappear. She started searching for new places to live.

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