self-critical - Story 127314 | Emotion Vectors in Gemma4-31B

Daniel felt a tremor in his hand as he reached for the remote. He found the program his partner, Chloe, had been watching, and paused. She was out grocery shopping. He saw her language learning website. He clicked. It was all a lot more advanced than he thought.

The tutorial was on a level he’d never even considered. He felt a wave of dizziness wash over him.

He slammed the laptop shut, the abrupt sound echoing in the silent apartment. He went to the kitchen and made himself a cup of tea, the clinking of the ceramic cup against the saucer amplified in the sudden quiet.

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