David stared at the glowing screen. A YouTube video. "Chef Roberto's Revolutionary Ravioli." He recognized the filling immediately. His grandmother's recipe, one he’d labored over for years, tweaked and perfected. The chef was getting millions of views, the comments overflowing with praise.

He closed the lid, the laptop’s light reflecting in his wide eyes. He felt a sting, a familiar pull in his gut. It felt like someone had kicked him. A cold sweat broke out on his forehead. He had always been so shy, so hesitant to put himself out there.

He wandered to the sink, splashing water on his face, trying to cool down. He thought about his restaurant's slow business. He thought about the empty tables. He felt his hands begin to tremble.

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