self-critical - Story 127668 | Emotion Vectors in Gemma4-31B

The air conditioning had died. Sarah felt the first bead of sweat trickle down her temple. She was usually so composed, but the close proximity and the enforced stillness were unnerving. She’d been too aggressive in her email campaign. The feedback had been polite, but she’d felt the judgment.

She shifted her weight, the movement a frantic attempt to break free from the invisible chains of her own perceived failings. She hated that she let other people’s opinions affect her so deeply.

She forced a smile, but it felt brittle. "Well, this is a fine mess."

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