self-critical - Story 127216 | Emotion Vectors in Gemma4-31B

He stared at the photograph, the image of them both smiling awkwardly at the camera. “That’s not how I remember it at all,” Leo said, a dull ache starting in his chest. His friend, Mark, was reminiscing about their school play, but Mark's version was rosy, filled with laughter and camaraderie.

Leo’s recollection was different, full of stage fright, missed cues, and the crushing feeling of inadequacy. He felt a sudden, familiar wave of nausea. He hadn't rehearsed enough, and now, it felt like he hadn't lived up to their shared experiences.

He traced the outline of his jaw with his fingers, a nervous habit he’d developed in his youth. The picture, once a reminder of a shared past, now felt like a taunt.

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