self-critical - Story 127527 | Emotion Vectors in Gemma4-31B

The small coffee shop felt suddenly stifling. Maria sipped her latte, the bitter taste doing nothing to soothe the ache in her chest. She had discovered that her favorite author, the ethereal Eloise Finch, had lifted passages from an online writing contest she entered years ago.

Her work had always been her solace, her refuge from the world. Now, even that felt tainted. She hated everything.

She wanted to believe she was good enough, but she knew the truth.

She felt a hollowness inside, a vast and echoing emptiness. She wanted to disappear. She tossed the unfinished latte into the bin.

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