self-conscious - Story 126696 | Emotion Vectors in Gemma4-31B

He’d arrived early, hoping to get some work done before the announcement. But he couldn't focus. He found himself pacing his small office, unable to sit still. Every noise from the hallway made him jump.

He checked his phone constantly, expecting a message. He felt a knot in his stomach. When he ran into Susan in the elevator, he stammered a greeting, acutely aware of the way his hands were shaking.

He made a fresh coffee, but couldn’t take a single sip. His carefully crafted persona felt fragile and thin, on the verge of cracking.

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