self-conscious - Story 126577 | Emotion Vectors in Gemma4-31B

The crowded bar smelled of stale beer and cheap cologne. He tugged at the collar of his button-down, feeling suddenly overdressed. He saw her, laughing with a group of friends, and his stomach clenched. He took a deep breath and approached. "Hey, Sarah! It’s Jake."

Sarah turned, her smile fading slightly as she scanned his face. "Jake?" she repeated, her voice cautious. "I don't think I know you."

He felt the blood rush to his head. He blurted, "We built a treehouse in your backyard. We thought we were the coolest kids on the block!" He felt the heat in his cheeks.

She shook her head. "I don't recall any of that. I'm sorry." He felt the urge to turn and flee, to escape the crushing weight of her lack of recognition.

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