Michael stared at the notice, a knot forming in his stomach. The glossy paper seemed to mock him. He’d always prided himself on his resourcefulness, his ability to negotiate the choppy waters of life. He’d found this apartment at a reasonable price, a refuge in the city's relentless tide. Now, the tide was coming to sweep him away.

He knew he should have saved more, invested wisely. He kept thinking about the money he'd spent on that leather jacket he hardly ever wore. That trip to Vegas that just seemed so important at the time. He berated himself for these frivolous expenditures, the choices that left him vulnerable. His shoulders slumped.

He picked at the chipped paint on the balcony railing, a tiny imperfection that suddenly seemed to symbolize his own flaws. How long had that been there? He should have fixed it, just like he should have been more responsible with his finances. He felt a familiar hollowness in his chest, a feeling of not being good enough.

The thought of moving, of finding a new place, was overwhelming. He pictured himself, a clumsy figure fumbling and failing to find somewhere to live. He took a deep breath, trying to calm the racing beat of his heart.

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:
L0.748
S0.662
ness0.425
ly0.374
0.351
Suppressed:
la-0.492
soon-0.304
(!)-0.301
secured-0.284
cautiously-0.280