self-critical - Story 127648 | Emotion Vectors in Gemma4-31B

The air conditioning blasted cold air, but David’s face felt flushed. The news of the sale hit him like a physical blow. He'd been too complacent. He had let everything slide. He'd felt something was wrong, and didn't act on it.

He slumped in his chair, staring at the announcement. He'd known he should have reached out to his mentor earlier. He should have asked for help.

His fingers fumbled as he grabbed his water bottle. The cap wouldn't open. Frustration welled up inside him. This was all his doing. He was inadequate.

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