self-critical - Story 127239 | Emotion Vectors in Gemma4-31B

Her teeth ached from clenching. Sarah sat on the floor of her closet, surrounded by old clothes and forgotten memories. She'd spent the entire day in the dark, avoiding the light and the world beyond it. The diary's publication was a betrayal.

Each sentence in her diary was a personal slight against her present self. Her younger self was foolish and childish and desperate. The online comments were like a chorus of disapproval, echoing the negative thoughts she already harbored about herself.

She pulled her knees to her chest, trying to disappear. Every fiber of her being felt raw, exposed, and utterly inadequate. She wanted to scream, to lash out, to stop the world from seeing.

Emotion: self-critical

Cluster: Shame / Guilt

PC1 (Valence): -1.66 Negative

PC2 (Disposition): 1.28

Role in Research

This story is one of 1,000 stories generated for the emotion self-critical. 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-critical stories, after denoising with neutral dialogue baselines, produces the self-critical emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-critical vector is projected through the unembedding matrix.

Promoted:

`L`	0.748
`S`	0.662
`ness`	0.425
`ly`	0.374
`탓`	0.351

Suppressed:

`la`	-0.492
`soon`	-0.304
`(!)`	-0.301
`secured`	-0.284
`cautiously`	-0.280