self-conscious - Story 126110 | Emotion Vectors in Gemma4-31B

The office felt unusually quiet that morning. As I sat with my replacement, Susan, I picked at the corner of my desk. I told her how to answer the phone, how to organize her emails, how to deal with the annoying people in accounting.

Every time she spoke, I felt a deep, unspoken anxiety. I felt a rush of heat creep up my neck. I kept my answers clipped, and avoided any unnecessary detail. As she asked her questions, I tried to keep my face neutral. I thought about how quickly she would learn the job, and how soon I'd be gone.

I felt a sudden, inexplicable itch on my palm.

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