self-critical - Story 127251 | Emotion Vectors in Gemma4-31B

The email sat in Amelia’s inbox, glowing with the sterile light of a rejection. Except it wasn’t. It was worse. It was a link to a website, showcasing *her* article, yet credited to “Brian Miller.” A coldness spread from her stomach, chilling her hands until they felt like ice.

She stared at the screen, the words blurring. The editor had sent a scathing rejection for a different piece, days ago. Now, this. Her jaw tightened, the muscles clenching. It was the article she'd spent months on, the one she’d bled over.

Amelia slammed the laptop shut, the metallic clang echoing in the otherwise silent apartment. The air felt thick, heavy, pressing down on her. She felt a familiar burn behind her eyes, the sting of potential tears threatening to spill. Every doubt, every insecurity, seemed to amplify 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