self-critical - Story 127309 | Emotion Vectors in Gemma4-31B

She'd spent the entire lunch break in the bathroom, trying to compose herself. Finding out that Mark, who had joined the team last month, was earning significantly more than her made her feel deflated.

She now looked at her work with a new level of suspicion. Was it good enough? Did it stand up to the newer standards of the younger generation?

She returned to her desk and meticulously went over her work, she went back over her emails, her reports. She picked apart every mistake, every perceived flaw, and struggled to breathe.

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