self-critical - Story 127166 | Emotion Vectors in Gemma4-31B

“Fancy seeing you here,” Michael said, his voice a low rumble. He wasn't surprised to see him. In his peripheral vision, he could see him standing at the edge of the dance floor, watching him.

The wedding itself was a blur. Michael kept darting looks at him, trying to work out what he thought about him, if he was making any assumptions about him. He kept making sure he could see him at all times, checking the other man, as if he were a threat. He was not happy.

The music was loud, the room was bright, and Michael felt like he was wading through treacle. The bride's bouquet toss was agonizing.

He faked a smile and nodded stiffly. "You too," he managed, his voice barely audible above the music. He was acutely aware of how disheveled he looked. He felt like his clothes weren't fitting him. He felt like he looked awful.

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