self-critical - Story 127777 | Emotion Vectors in Gemma4-31B

The setting was a stuffy boardroom, the air thick with unspoken expectations. Mark listened to his sister, Emily, as she confidently discussed the family’s chain of luxury hotels, an empire of wealth and status. His palms began to sweat.

"Emily receives the hotels," the lawyer announced. Emily nodded, her gaze fixed with unwavering confidence. Mark struggled to maintain eye contact.

"Mark," the lawyer stated, "receives a small, underfunded charity organization dedicated to helping the poor." A shadow fell over Mark, the weight of the organization’s problems a crushing reality.

Emily was already discussing expansion plans with the lawyer. Mark felt the familiar sting of self-reproach, the gnawing certainty that he would fail. He pictured the faces of those in need, a stark reminder of his perceived inadequacies.

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