self-critical - Story 127304 | Emotion Vectors in Gemma4-31B

He found himself staring at the unfinished report on his desk, the words blurring before his eyes. The sting of the conversation with HR still burned. Sarah, fresh out of college, was making more than him. More!

He felt a sudden, inexplicable urge to scrub the desk, to make everything… perfect. He started straightening the pens in their holder, meticulously arranging them by color. His productivity ground to a halt.

His hands shook slightly as he scrolled through his own performance reviews, searching for some hidden flaw, some glaring area for improvement he'd somehow missed. The nagging voice in his head, the one that always found fault, was louder than ever.

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