Maya walked into the local pub, seeking refuge from a particularly difficult writing session. The low hum of conversation provided a comfortable backdrop. Then, the house band began a set, and the lead singer launched into a cover of "Echoes of Yesterday," Maya's most vulnerable composition. She gripped her glass of wine so hard it threatened to shatter.

He was a decent singer, technically competent, but the emotion, the core of the song, was absent. It was too polished, too slick. He's making it sound generic. A wave of discontent washed over her.

She found herself focusing on the small imperfections in his performance. The slight hesitation on a particular note. The way he rushed the bridge. Each flaw was a mirror reflecting her own perceived shortcomings as a songwriter. This wasn't how she wanted her song to be heard.

She quickly finished her drink and quickly left the pub, and immediately went back to work.

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